39.2 COLUMBIA JOURNAL OF GENDER AND LAW                         91 

 

 

 

MISS DIAGNOSIS: GENDERED INJUSTICE IN 

MEDICAL MALPRACTICE LAW 
 

CECILIA PLAZA* 
 

Abstract 

 

Women patients have experienced a history of discrimination in medical practice. 

Medical malpractice litigation offers an avenue for patients affected by practitioner 

negligence to recoup the costs inflicted by their injuries. The present study investigates 

the impact of patient gender on plaintiffs’ recovery amounts in medical malpractice suits 

alleging delayed, wrongful, or misdiagnosis, as women are more vulnerable to diagnosis-

related malpractice. This study also analyzes the impact of contextual factors such as 

state demographics, state malpractice legislation, and features of each instance of 

litigation, such as the duration of each case. Using a national database of resolved 

malpractice cases from 2004 to 2018, this study uses several different statistical models 

to shed light on the contours of the gender gap in medical malpractice litigation. This 

study also offers suggestions for future research and potential solutions to address the 

gender gap and increase equal access to legal recourse after medical injury due to 

negligence for patients of all genders. 

 

INTRODUCTION 

 

Women have historically experienced disadvantages in accessing quality medical 

care.1 One major disadvantage that women face relative to their male counterparts is the 

risk of misdiagnosis—which includes delays in diagnosis, wrongful diagnosis, and no 

 
* The author would like to thank Professor Jeffrey Fagan of Columbia Law School for his guidance, 

supervision, and support throughout this study. The author would also like to thank Professor Kristen 

Underhill of Columbia Law School for her feedback and helpful insight. 

 
1 The author recognizes that concerns about quality of care apply to other minority groups, including gender 

minority and racial minority groups, and that gender intersects with other identity factors like race and age. 

Given the limitations of the data, see infra Part IV.A., the focus of this study is on women as identified by 

national datasets. See infra Part III.A. for more detailed information on the datasets used. See infra Parts 

I.C.4. and IV for further discussion of intersectional disadvantages for plaintiffs in medical malpractice suits 

and how to address such intersectionality in future research. 



92 COLUMBIA JOURNAL OF GENDER AND LAW 39.2  

diagnosis at all.2 Women are more likely to face misdiagnosis than men.3 This can be 

attributed to two main factors: the dearth of medical scientific knowledge about women’s 

health4 and the widespread distrust among health professionals of their women patients 

relative to male patients.5 The combination of the “knowledge gap”—the medical 

community’s lack of knowledge about women’s health due to women’s historical 

underrepresentation in medical research—and the trust gap—the medical profession’s 

history of distrusting or downplaying women’s reports of their own symptoms—creates 

an increased risk of missed, delayed, and incorrect diagnoses for women.6 

 

 

 
2 See Trisha Torrey, How Common Is Misdiagnosis or Missed Diagnosis?, VERY WELL HEALTH (May 30, 

2019), https://www.verywellhealth.com/how-common-is-misdiagnosis-or-missed-diagnosis-2615481 [https:// 

perma.cc/WT9L-G4ET] (defining a misdiagnosis as “a situation when your doctor tells you that you have 

some illness or condition, but it’s incorrect” and a missed diagnosis as “the lack of a diagnosis, usually 

leading to no or inaccurate treatment”). 

 
3 See generally David E. Newman-Toker et al., Missed Diagnosis of Stroke in the Emergency Department: A 

Cross-Sectional Analysis of a Large Population-Based Sample, 1 DIAGNOSIS 155, 158, 166 (2014) (finding 

that women are more likely than men to have their strokes misdiagnosed); Kiera Carter, The Frustrating 

Reasons Why Doctors Constantly Misdiagnose Women, PREVENTION (Jan. 31, 2019), https://www. 

prevention.com/health/a26100121/misdiagnosed-women/ [https:// perma.cc/RD4Q-S45V] (noting that 

women are more likely than men to be misdiagnosed when they are experiencing heart attacks, autoimmune 

diseases, or sex-specific conditions); Gina Shaw, Why Women Struggle to Get the Right Diagnosis, WEBMD 

HEALTH NEWS (June 8, 2018), https://www.webmd.com/ women/news/20180607/why-women-are-getting-

misdiagnosed [https://perma.cc/VBC8-FR8H] (noting that women are more likely than men to be 

misdiagnosed during a heart attack or stroke, are more likely to have their autoimmune diseases 

misdiagnosed, and more likely to have their pain underestimated by health professionals); Laura Kiesel, 

Women and Pain: Disparities in Experience and Treatment, HARVARD HEALTH BLOG (Oct. 9, 2017), 

https://www.health.harvard.edu/blog/women-and-pain-disparities-in-experience-and-treatment-

2017100912562 [https://perma.cc/UW7T-WX29] (noting that women’s pain is more likely than men’s to be 

underestimated by medical professionals). 

 
4 See, e.g., CAROLINE CRIADO PEREZ, INVISIBLE WOMEN: DATA BIAS IN A WORLD DESIGNED FOR MEN 198 

(2019) (referring to this phenomenon as the “gender data gap” and applying it to various aspects of life 

beyond medicine). See generally MAYA DUSENBERY, DOING HARM: THE TRUTH ABOUT HOW BAD MEDICINE 

AND LAZY SCIENCE LEAVE WOMEN DISMISSED, MISDIAGNOSED, AND SICK 53–137 (2018) [hereinafter 

DUSENBERY, DOING HARM]. 

 
5 See, e.g., DUSENBERY, DOING HARM, supra note 4, at 187–88 (discussing the history of medical 

professionals discounting women’s pain); see also supra note 3 and accompanying text. 

 
6 The terms “knowledge gap” and “trust gap” were coined by Maya Dusenbery. DUSENBERY, DOING HARM, 

supra note 4, at 23, 61. Caroline Criado Perez referred to the knowledge gap alternatively as the “gender data 

gap,” but applied the term more broadly to aspects of daily life beyond health and medicine. CRIADO PEREZ, 

supra note 4, at 198. 



 

39.2 COLUMBIA JOURNAL OF GENDER AND LAW                         93 

 

 

 

The legal realm of medical malpractice offers a remedy for individuals who have 

been adversely affected by medical errors due to negligence, including those who have 

suffered wrong, delayed, or missed diagnoses.7 However, the legal standard for medical 

malpractice currently protects the common practice of discounting women patients and 

their reports of their symptoms. The malpractice standard only requires that practitioners 

abide by what ordinary practitioners in their field would do in similar situations;8 the 

pervasiveness of the knowledge and trust gaps suggests that this standard more negatively 

impacts women who experience and subsequently file suit for diagnosis-related 

malpractice than their male counterparts. 

 

The present study analyzes the interaction of the knowledge and trust gaps with the 

legal standard for medical malpractice and the resulting impact on women’s diagnosis-

related medical malpractice lawsuits. Part I offers a brief history of the knowledge and 

trust gaps and their impact on patient care. Part II presents an overview of medical 

malpractice law and the gaps in the law which perpetuate gender bias. Part III outlines the 

methodology used for this study to assess the impact of the knowledge and trust gaps on 

women patients who pursue medical malpractice actions. Part III also explains various 

statistical models and their usefulness in explaining gender bias in medical malpractice 

actions. Part IV summarizes the results of this study and discusses the implications for 

women patients who pursue medical malpractice lawsuits. Part IV concludes by 

suggesting legal reforms to address gender bias in medical malpractice law.  

 

Ultimately, this study finds that patient gender does significantly impact medical 

malpractice lawsuit outcomes and that gender interacts with other key contextual factors 

to produce such disparities. The presence of a gendered impact indicates that women are 

disadvantaged both as patients in the healthcare setting and later as plaintiffs in the legal 

setting, and that the current legal standard for medical malpractice offers a disincentive to 

change medical practice and patient care. 

 

 

 

 

 

 
7 See generally BARRY R. FURROW ET AL., HEALTH LAW: CASES, MATERIALS AND PROBLEMS 207–47 (8th ed. 

2018) (discussing medical malpractice as a legal tool available to those patients who can show their medical 

practitioner fell below the standard of care that a medical professional is legally required to give to a patient). 

See infra Part II for more information about medical malpractice and the legal standard of care. 

 
8 See James S. Higgins, Defense of Medical Malpractice Cases, in 16 AM. JUR. 2D Evidence § 137 (1998). 



94 COLUMBIA JOURNAL OF GENDER AND LAW 39.2  

I. The Knowledge Gap, the Trust Gap, and the Consequences for Women 

 

A. The Knowledge Gap 

 

There is less information available about women’s health than about men’s health.9 

In other words, there is a knowledge gap between the medical profession’s understanding 

of men’s health and women’s health. This is a natural consequence of women’s historical 

underrepresentation in biomedical studies and the relative neglect of diseases that 

predominantly affect women in medical research, both of which continue to this day 

despite patient advocacy efforts.10 The resulting knowledge gap affects the quality of 

medical care that women patients receive.  

 

The male body as the default for medical science goes back to the Ancient Greeks: 

Aristotle believed women’s bodies were simply “mutilated” male bodies.11 “Ovaries” 

were referred to as female testicles and were not named as a separate organ until the 

seventeenth century.12 Even in the twenty-first century, the male body is still often 

depicted as the default human body; medical textbooks often use illustrations of the male 

body to refer to neutral body parts. A 2008 study of over 16,000 images in textbooks 

recommended by prestigious universities in the United States, Europe, and Canada found 

that male bodies were used three times more frequently than female bodies to illustrate 

non-reproductive body parts.13 This is not a harmless error; there are actually sex-based 

 

 
9 See CRIADO PEREZ, supra note 4, at 196 (explaining that medical science research has historically been 

based on the “average” male body); Reshma Jagsi et al., Under-Representation of Women in High-Impact 

Published Clinical Cancer Research, 115 CANCER 3293, 3299 (2009) (finding that women are 

underrepresented in cancer research); Sherry A. Marts & Sarah Keitt, Foreword: A Historical Overview of 

Advocacy for Research in Sex-Based Biology, 34 ADVANCES IN MOLECULAR & CELL BIOLOGY v (2004) 

(discussing a long-standing need to collect more data on women’s health and bodies). 

 
10 See DUSENBERY, DOING HARM, supra note 4, at 24–25 (noting that women have been excluded from many 

large-scale, seminal studies like the Baltimore Longitudinal Study of Aging, the Physician’s Health Study, 

the 1982 Multiple Risk Factor Intervention Trial, and even some major studies on breast cancer). 

 
11 See CRIADO PEREZ, supra note 4, at 196. 

 
12 See id. 

 
13 Plataforma SINC, Medical Textbooks Use White, Heterosexual Men As a ‘Universal Model’, SCIENCE 

DAILY (Oct. 17, 2008), https://www.sciencedaily.com/releases/2008/10/081015132108.htm [https://perma. 

cc/PTY6-58UU]. 



 

39.2 COLUMBIA JOURNAL OF GENDER AND LAW                         95 

 

 

 

differences in human tissues, organs, and the course and impact of diseases.14 Yet women 

have been and continue to be underrepresented in medical research.15 Thus, for centuries, 

medical professionals have been treating women patients according to a male model that 

often does not fit women’s bodies. 

 

The National Institutes of Health (NIH), the largest public funder of biomedical 

research,16 took note of this issue in 1985, when it issued new report on gender 

representation in medical research.17 But not much tangible change occurred until after 

1990, when the Society for the Advancement of Women’s Health Research (SWHR) 

demanded an audit of the NIH by the U.S. Government Accountability Office (GAO).18 

Following the GAO audit, the NIH formed its Office of Research on Women’s Health 

(ORWH);19 mandated the inclusion of women as research subjects; and launched the 

Women’s Health Initiative, a huge medical research study to collect the data on women 

that was already available for men.20 In 1993, the NIH Revitalization Act made the 

ORWH a permanent part of the NIH21 and mandated that NIH-funded studies include 

 
14 See Natasha A. Karp et al., Prevalence of Sexual Dimorphism in Mammalian Phenotypic Traits, 8 NATURE 

COMM., June 26, 2017, https://www.nature.com/articles/ncomms15475 [https://perma.cc/9BB8-8KLM]. 

 
15 See Jagsi et al., supra note 9, at 3299 (finding that women are underrepresented in cancer research); see 

also Press Release, George Washington University Public Health, Women Remain Underrepresented in 

Medical Science, New Report Says (Mar. 3, 2014) (on file with George Washington University), https:// 

publichealth.gwu.edu/content/women-remain-underrepresented-medical-science-new-report-says [https:// 

perma.cc/PEM8-U6HL] (“Women are still underrepresented in biomedical research, and when included, 

often the differences between women and men are not analyzed fully. This is a gap that must be addressed in 

order to achieve better outcomes for women.”). 

 
16 NAT’L INSTS. OF HEALTH, OUR SOCIETY, https://www.nih.gov/about-nih/what-we-do/impact-nih-

research/our-society [https://perma.cc/9FBC-M3E6]. 

 
17 See generally U.S. PUB. HEALTH SERV., PUB. HEALTH REPORTS 100, NO. 1, REPORT OF THE PUBLIC HEALTH 

SERVICE TASK FORCE ON WOMEN’S HEALTH ISSUES (1985). 

 
18 See U.S. GOV’T ACCOUNTABILITY OFFICE, GAO/T-HRD-90-38, NATIONAL INSTITUTES OF HEALTH: 

PROBLEMS IN IMPLEMENTING POLICY ON WOMEN IN STUDY POPULATIONS, TESTIMONY BY MARK V. VADEL 

BEFORE SUBCOMMITTEE ON HEALTH AND THE ENVIRONMENT, COMMITTEE ON ENERGY AND COMMERCE, HOUSE 

OF REPRESENTATIVES (1990). 

 
19 See U.S. DEP’T HEALTH & HUMAN SERV., 92-3457, REPORT OF THE NATIONAL INSTITUTES OF HEALTH: 

SEPTEMBER 4–6, 1991 (1992). 

 
20 See LESLIE LAURENCE & BETH WEINHOUSE, OUTRAGEOUS PRACTICES: HOW GENDER BIAS THREATENS 

WOMEN’S HEALTH 78–79 (1994). 

 
21 42 U.S.C. § 287d (2018). 



96 COLUMBIA JOURNAL OF GENDER AND LAW 39.2  

enough women and racial minorities to conduct a valid analysis of differences.22 It also 

mandated increased efforts to research conditions that disproportionately affect women, 

such as breast and ovarian cancers and osteoporosis.23 While this was a sign of much-

needed improvement, these measures were not as progressive as they seemed: The 

mandate on women’s inclusion only applied to later phases of research and studies 

enrolling human participants, so studies at the cellular level or in the animal research 

phase could still be all male, an imbalance not addressed by the NIH until 2014.24 These 

measures also failed to have the desired impact. Women continue to be underrepresented 

in medical research,25 and even if adequately represented, the majority of studies that 

include women don’t conduct or publish gender difference analyses.26 Research on 

diseases disproportionately affecting women also remain underfunded compared to 

diseases that predominantly affect men.27  

 

 

 
22 See Guideline for the Study and Evaluation of Gender Differences in the Clinical Evaluation of Drugs, 58 

Fed. Reg. 39,406 (July 22, 1993). 

 
23 42 U.S.C. § 284e (2018). 

 
24 See Janine A. Clayton & Francis S. Collins, Policy: NIH to Balance Sex in Cell and Animal Studies, 509 

NATURE 282, 282 (2014). 

 
25 A 2000 GAO audit revealed individual studies still weren’t analyzing results by gender. U.S. GOV’T 

ACCOUNTABILITY OFFICE, GAO/HEHS-00-96, NIH HAS INCREASED ITS EFFORTS TO INCLUDE WOMEN IN 

RESEARCH (2000). A 2015 GAO review stated that it was difficult to gauge improvement in gender 

representation from the NIH’s aggregate data. U.S. GOV’T ACCOUNTABILITY OFFICE, GAO-01-754, BETTER 

OVERSIGHT NEEDED TO HELP ENSURE CONTINUED PROGRESS INCLUDING WOMEN IN HEALTH RESEARCH 

(2015). Independent studies show women have remained significantly underrepresented. See, e.g., Carolyn 

M. Mazure & Daniel P. Jones, Twenty Years and Still Counting: Including Women as Participants and 

Studying Sex and Gender in Biomedical Research, 15 BOS. MED. CTR. WOMEN’S HEALTH 94 (2015); Stacie E. 

Geller et al., Inclusion, Analysis, and Reporting of Sex and Race/Ethnicity in Clinical Trials: Have We Made 

Progress?, 20 J. WOMEN’S HEALTH 315 (2011). 

 
26 See Basmah Safdar et al., Inclusion of Gender in Emergency Medicine Research, 18 ACAD. EMERGENCY 

MED. e1, e3 (2011) (finding that few emergency medicine studies examine the effect of gender on health 

outcomes); Andrea H. Weinberger, Sherry A. McKee & Carolyn M. Mazure, Inclusion of Women and 

Gender-Specific Analyses in Randomized Clinical Trials of Treatments for Depression, 19 J. WOMEN’S 

HEALTH 1727, 1730–31 (2010) (finding that, while understanding gender differences is crucial to effective 

treatment, many recent studies of depression treatment do not examine outcomes by gender). 

 
27 See Anita Holdcroft, Gender Bias in Research: How Does it Affect Evidence Based Medicine?, 100 J. 

ROYAL SOC’Y MED. 2, 2 (2007) (noting that even when research includes women, sex-specific differences are 

not studied); Ruth L. Kirschstein, Research on Women’s Health, 81 AM. J. PUB. HEALTH 291, 292 (1991) 

(expressing the need to continue increasing funding for women’s health research). 



 

39.2 COLUMBIA JOURNAL OF GENDER AND LAW                         97 

 

 

 

Moreover, while it has received the most legislative attention regarding women’s 

representation, the NIH is not the only funder of biomedical research. Most research is 

privately funded by pharmaceutical companies and reviewed by the U.S. Food and Drug 

Administration (FDA).28 There are no federal guidelines requiring these private funders 

to include women in their studies and the FDA has not implemented any such 

requirements into its approval process. In fact, from 1977 to 1993, the FDA explicitly 

prohibited women “of childbearing potential” from participating in early phase drug 

trials,29 which had a chilling effect on women’s representation altogether; researchers 

were hesitant to enroll women even in later phases of research and women remain 

underrepresented in the majority of late phase studies.30 The FDA still does not require 

representation of female cells or animals in early phase research nor gender difference 

analyses for late phase research.31 As a result, women have been left out of some of the 

biggest and most influential medical studies in recent history, and continue to be 

underrepresented in biomedical research.32 For instance, women were completely left out 

of the Baltimore Longitudinal Study of Aging, which purported to explore “normal 

human aging;” the Physician’s Health Study, also known as the famous ‘baby aspirin’ 

study on the preventative effects of aspirin on heart disease; the 1982 Multiple Risk 

Factor Intervention Trial studying the effects of diet and exercise on heart disease; and 

even studies examining the effects of obesity on breast cancer.33 

 

That is not to say that medical knowledge about women’s health has not advanced. 

However, the knowledge we do have about women’s health has not yet translated well 

into medical education and the practice of medicine, and our medical models are still 

 
28 See Earl Ray Dorsey et al., Funding of US Biomedical Research, 2003-2008, 303 JAMA 137, 140 (2010) 

(identifying private biopharmaceutical companies as the biggest contributors to modern medical research). 

 
29 See U.S. FOOD & DRUG ADMIN., GENDER STUDIES IN PRODUCT DEVELOPMENT: HISTORICAL OVERVIEW, 

https://www.fda.gov/science-research/womens-health-research/gender-studies-product-development-

historical-overview [https://perma.cc/8A5D-EXYG]. 

 
30 See U.S. GOV’T ACCOUNTABILITY OFFICE, GAO/HRD-93-17, FDA NEEDS TO ENSURE MORE STUDY OF 

GENDER DIFFERENCES IN PRESCRIPTION DRUGS TESTING (1992). 

 
31 See U.S. GOV’T ACCOUNTABILITY OFFICE, GAO-01-754, WOMEN SUFFICIENTLY REPRESENTED IN NEW 

DRUG TESTING, BUT FDA OVERSIGHT NEEDS IMPROVEMENT (2001). 

 
32 See Kat Kwiatkowski et al., Inclusion of Minorities and Women in Cancer Clinical Trials, a Decade Later: 

Have We Improved? 119 CANCER 2956, 2961–62 (2013) (finding that women continue to be 

underrepresented in clinical trials). 

 
33 DUSENBERY, DOING HARM, supra note 4, at 25. 



98 COLUMBIA JOURNAL OF GENDER AND LAW 39.2  

based on the average male body as a default.34 The first United States textbook on 

gender-specific medicine was not published until 2004.35 By 2011, about 70% of medical 

schools in the United States and Canada still had minimal to no curriculum coverage on 

sex and gender differences.36 

 

This history of women’s exclusion has led to a significant knowledge gap, meaning 

the medical community has less information about women’s bodies and health than it 

does about men’s bodies and health. This has obvious implications for the quality of 

women’s healthcare. But there is more to the story than the knowledge gap; there is also a 

deep history of distrust of women patients that impacts quality of care. 

 

B. The Trust Gap 

 

The trust gap is the increased distrust that medical professionals have historically had 

and still have of women patients as compared to men. In general, medical professionals 

are more likely to discount women patient’s symptoms in favor of their own knowledge 

about what illnesses ‘typically’ affect women.37 Professionals are also more likely to 

discount women’s symptoms altogether and label them as psychological in origin.38 This 

leads to a pervasive distrust between women and their medical practitioners: 

Professionals do not trust women patients as reliable reporters of their own symptoms 

and, as women patients feel discounted and distrusted, they in turn distrust their medical 

professionals. 

 

 

 
34 See U.S. DEP’T HEALTH & HUMAN SERV., HRSA-P-DM-95-1, FIFTH REPORT: WOMEN AND MEDICINE 

(1995). 

 
35 MARIANNE J. LEGATO, PRINCIPLES OF GENDER-SPECIFIC MEDICINE (2004). 

 
36 See Virginia M. Miller et al., Embedding Concepts of Sex and Gender Health Differences into Medical 

Curricula, 22 J. WOMEN’S HEALTH 194, 196 (2013). 

 
37 See generally Elizabeth M. Marks & Myra S. Hunter, Medically Unexplained Symptoms: An Acceptable 

Term?, 9 BR. J. PAIN 109, 111 (2015); Allen Frances, The New Somatic Symptom Disorder in DSM-5 Risks 

Mislabeling Many People as Mentally Ill, 346 BR. MED. J. f1580, f1581 (2013); David Edelberg, 

Fibromyalgia Confounds Allopathic Habits of Mind, 14 AMA J. ETHICS 305, 306–07 (2012); AM. 

PSYCHIATRIC ASS’N, DIAGNOSTIC AND STATISTICAL MANUAL OF MENTAL DISORDERS 309–29 (5th ed. 2013); 

U.S. DEP’T HEALTH & HUMAN SERV., supra note 34. 

 
38 See Marks & Hunter, supra note 37, at 111; Frances, supra note 37, at f1581; Edelberg, supra note 37, at 

306–07; AM. PSYCHIATRIC ASS’N, supra note 37, at 309–29. 



 

39.2 COLUMBIA JOURNAL OF GENDER AND LAW                         99 

 

 

 

This history of distrust begins with hysteria in the fifth century, when the term was 

first used to refer to illnesses in women caused by the womb—which, at the time, was 

nearly every illness that could afflict a woman.39 In the eighteenth century, this definition 

changed such that hysteria became a catch-all category for women’s illnesses, 

particularly those thought to be psychiatric in origin.40 This idea of a connection between 

women, their reproductive organs, and psychiatric illnesses stuck around through the 

nineteenth century41 and was solidified in the twentieth century with the Freudian school 

of thought.42 The significance of the history of hysteria for modern women is that the idea 

that women’s illnesses and symptoms are tied to women’s reproductive systems—or the 

“bikini medicine” approach43—and the idea that women’s symptoms are more likely than 

men’s to be psychological in origin are deeply ingrained in the history of the medical 

profession. Symptoms or illnesses that cannot be attributed to something under the 

“bikini” umbrella are likely to be categorized as “somatoform disorders,” psychogenic 

illnesses, or “medically unexplained symptoms” (MUS), all of which are modern 

permutations of the age-old category of hysteria, used to imply that the patient’s 

symptoms are psychological or otherwise ‘not real.’44  

 

The Diagnostic and Statistical Manual of Mental Disorders (DSM) describes these 

psychogenic illnesses as more common among women than men45 and does not contain 

even a warning about ruling out other diagnoses before assigning a psychogenic 

explanation to a patient’s symptoms, which Dr. Allen Frances, chair of the task force on 

the fourth edition of the DSM, warned could lead to widespread misdiagnoses—

especially among women.46 Dr. Frances was rightfully concerned. Psychogenic 

 
39 See Helen King, Once Upon a Text: Hysteria from Hippocrates, in HYSTERIA BEYOND FREUD 3, 12–13 

(Sander L. Gilman et al. eds., 1993). 

 
40 See RICHARD BLACKMORE, A TREATISE OF THE SPLEEN AND VAPOURS: OR, HYPOCHONDRIACAL AND 

HYSTERICAL AFFECTIONS 96–114 (1726). 

 
41 See generally Paul Chodoff, Hysteria and Women, 139 AM. J. PSYCHIATRY 546 (1982). 

 
42 See CARROLL SMITH-ROSENBERG, DISORDERLY CONDUCT: VISIONS OF GENDER IN VICTORIAN AMERICA 197 

(1986). 

 
43 See Nanette K. Wenger, You’ve Come a Long Way, Baby: Cardiovascular Health and Disease in Women: 

Problems and Prospects, 109 CIRCULATION 558, 558 (2004). 

 
44 See Marks & Hunter, supra note 37, at 111; Edelberg, supra note 37, at 306–07. 

 
45 AM. PSYCHIATRIC ASS’N, supra note 37, at 309–29. 

 
46 See Frances, supra note 37, at f1581. 



100 COLUMBIA JOURNAL OF GENDER AND LAW 39.2  

explanations have been called the “wastepaper basket of medicine”47 because they are 

easy to use, they lack an objective test, and they are the only mental health diagnoses 

based not on symptomology but rather on speculation about cause.48 Even more 

concerning is how quickly healthcare professionals can jump to such diagnoses; in one 

study, that was as quickly as thirty seconds to two minutes into an interaction with a 

patient, depending on whether the physician felt confused, irritated at the patient, or felt 

that the interaction with the patient was negative.49 These psychogenic diagnoses become 

very ‘sticky:’ They shatter the credibility of a patient among other physicians because 

their medical file now says their symptoms are psychogenic.50 

 

Unsurprisingly, this affects women’s trust of medical professionals and the healthcare 

system.51 Stories of women’s unfortunate encounters with the healthcare system abound: 

It took Jackie ten years and many trips to specialists and the emergency room to finally 

be diagnosed with lupus.52 It took Jen Brea, a young Harvard PhD student, numerous 

visits to rheumatologists, neurologists, psychiatrists, and various other specialists; a 

faulty diagnosis of conversion disorder—a psychogenic explanation for physical 

symptoms; and four years before she was finally diagnosed with myalgic encephalitis, 

 
47 See ANGELA KENNEDY, AUTHORS OF OUR OWN MISFORTUNE? THE PROBLEMS WITH PSYCHOGENIC 

EXPLANATIONS FOR PHYSICAL ILLNESSES 17 (2012). 

 
48 See generally Laurie Endicott Thomas, Are Your Patient’s Medically Unexplained Symptoms Really ‘All in 

Her Head’?, 78 MED. HYPOTHESES 542 (2012). 

 
49 See Chaichana Nimnuan, Matthew Hotopf & Simon Wessely, Medically Unexplained Symptoms: How 

Often and Why Are They Missed?, 93 Q.J. MED. 21, 23, 25 (2000). 

 
50 See, e.g., Diane O’ Leary, Re: The New Somatic Symptom Disorder in DSM-5 Risks Mislabeling Many 

People as Mentally Ill, 346 BR. MED. J. f1580, f1580 (2013) (“A false positive diagnosis of somatic symptom 

disorder harms patients because it . . . subjects patients to stigma, inappropriate drugs, psychotherapy, and 

iatrogenic disease; disadvantages them in decisions relating to employment, education, and healthcare 

entitlements; skews their self perceptions [sic] and those of family and friends.”); AM. AUTOIMMUNE 

RELATED DISEASES ASS’N, WOMEN & AUTOIMMUNITY, https://www.aarda.org/who-we-help/patients/women-

and-autoimmunity/#1481574903922-68688035-6be6 [https://perma.cc/ZJ84-G7G4] (“Over 45% of 

autoimmune disease patients are labeled as ‘chronic complainers’ in early stages of their illness.”). 

 
51 See Annie W. Lin et al., Trust in Physicians and Medical Experience Beliefs Differ Between Women With 

and Without Polycystic Ovary Syndrome, 2 J. ENDOCRINE SOC’Y 1001, 1001 (finding that women with 

polycystic ovary syndrome were significantly more distrustful of their primary care physicians than the 

general population). 

 
52 See Maya Dusenbery, ‘Everybody Was Telling Me There Was Nothing Wrong’, BBC FUTURE (May 29, 

2018), https://www.bbc.com/future/article/20180523-how-gender-bias-affects-your-healthcare 

[https://perma.cc/EHB4-CKU7]. 



 

39.2 COLUMBIA JOURNAL OF GENDER AND LAW                         101 

 

 

 

more popularly known as chronic fatigue syndrome.53 Gila Lyons was told she was just 

suffering from anxiety before discovering that she had had a silverfish in her inner ear.54 

An unnamed woman with scleroderma was told her symptoms were “all in her head” 

before she was diagnosed, but not before she sustained such severe esophageal damage 

that she will never be able to eat again.55 Another woman with ovarian cancer was told 

for three years that she was experiencing early menopause.56 The list goes one. These 

women and their stories are the legacy of hysteria. 

 

Women are still more vulnerable to gender-biased diagnosing and “medical 

psychologizing” than men.57 They are more likely to receive psychogenic diagnoses58 and 

to have their physical symptoms attributed to stress or anxiety,59 even in the absence of 

any evidence of psychological illness or distress.60 This, combined with the stereotype 

that women tend to seek medical attention for minor or frivolous concerns—a stereotype 

 
53 TED, What Happens When You Have a Disease Doctors Can’t Diagnose, YOUTUBE (Jan. 17, 2017), 

https://www.youtube.com/watch?time_continue=1027&v=Fb3yp4uJhq0 [https://perma.cc/UZ5F-R659]. 

 
54 Gila Lyons, An Inch-Long Bug Lived in My Ear for Months, but My Doctor Dismissed it as Anxiety, OPRAH 

MAG. (Nov. 28, 2018), https://www.oprahmag.com/life/health/a25251599/woman-bug-in-ear-anxiety-

symptoms/ [https://perma.cc/SHU9-5T52]. 

 
55 TED, supra note 53. 

 
56 Id. 

 
57 See generally Shari Munch, Gender-Biased Diagnosing of Women’s Medical Complaints: Contributions of 

Feminist Thought, 1970–1995, 40 WOMEN & HEALTH 101 (2004) (describing the recent history of gender-

biased diagnosing); see also Karen Armitage, Lawrence J. Schneiderman & Robert A. Bass, Response of 

Physicians to Medical Complaints in Men and Women, 241 J. AM. MED. ASSOC. 2186, 2186 (1979) 

(concluding that the “data tend to support the argument that male physicians take medical illness more 

seriously in men than in women”). 

 
58 See Susan Wendell, Old Women Out of Control: Some Thoughts on Aging, Ethics, and Psychosomatic 

Medicine, in MOTHER TIME: WOMEN, AGING, AND ETHICS 133, 139–40 (Margaret U. Walker ed., 1999) 

(“[M]edically unexplained physical symptoms[] . . . [are] observed most frequently among women, 

individuals from non-Western or developing nations, [and] those from lower socioeconomic backgrounds.”). 

 
59 See Nancy N. Maserejian et al., Disparities in Physicians’ Interpretations of Heart Disease Symptoms by 

Patient Gender: Results of a Video Vignette Factorial Experiment, 18 J. WOMEN’S HEALTH 1661, 1661 

(2009) (finding that physicians frequently and confidently diagnosed middle-aged women with psychological 

conditions and middle-aged men with heart disease based on the same symptomology). 

 
60 See JEROME GROOPMAN, HOW DOCTORS THINK 264–65 (2007) (explaining the cognitive biases that might 

prompt a physician to tell a patient nothing is wrong with them and the damage such a statement might do, 

given the stigma of psychological illness). 



102 COLUMBIA JOURNAL OF GENDER AND LAW 39.2  

which has been disproven by research61—has led to the systematic discounting of 

women’s symptoms and a pervasive distrust of women patients by their physicians. 

Medical students are still taught through medical school curricula, both implicitly and 

explicitly, that women are unreliable reporters of their own symptoms and are more likely 

to have psychological rather than “real,” physical diseases.62 Emergency room physicians 

are still trained on the job to “be on the look-out for hysterical females.”63 The trust gap is 

still alive and well in clinics and hospitals around the country, discouraging women from 

seeking medical attention for fear of being discounted and discouraging practitioners 

from taking their women patients seriously. 

 

C. The Impact of Knowledge and Trust Gaps on Women’s Quality of Care 

 

When women’s symptoms are not taken seriously and when women patients are 

subjected to often incorrect psychogenic explanations of their symptoms,64 women 

patients as a group are left vulnerable to misdiagnosis, under-diagnosis, and to being 

stereotyped as mentally ill, making them less likely to ever receive a correct diagnosis.65 

Women patients are thus more likely to experience missed, delayed, or wrongful 

diagnoses, which are legally redressable claims under medical malpractice laws. 

 

 

 

 

 

 

 
61 See Kate Hunt et al., Do Women Consult More Than Men? A Review of Gender and Consultation for Back 

Pain and Headache, 16 J. HEALTH SERVS. RES. & POL’Y 108, 108 (2011) (finding that evidence of women 

consulting medical professionals more frequently than men was “weak and inconsistent”). 

 
62 As the first female dean of Harvard Medical School, Dr. Mary C. Howell, once noted: “[I]t is widely 

taught, both explicitly and implicitly, that women patients . . . are unreliable historians, and are beset by such 

emotionality that their symptoms are unlikely to reflect ‘real’ disease.” Mary C. Howell, What Medical 

Schools Teach About Women, 291 NEW ENGLAND J. MED. 304, 304 (1974). 

 
63 Denise Dador, ‘Medical Sexism’: Women’s Heart Disease Symptoms Often Dismissed, ABC EYEWITNESS 

NEWS (Nov. 2, 2011), http://abc7.com/archive/8416664/ (last visited Nov. 23, 2019) (quoting cardiologist Dr. 

Adam Splaver saying, “In training, we were taught to be on the lookout for hysterical females who come to 

the emergency room”). 

 
64 See generally Bruce L. Miller et al., Misdiagnosis of Hysteria, 34 AM. FAM. PHYSICIAN 157 (1986). 

 
65 See ELIZABETH A. KLONOFF & HOPE LANDRINE, PREVENTING MISDIAGNOSIS OF WOMEN: A GUIDE TO 

PHYSICAL DISORDERS THAT HAVE PSYCHIATRIC SYMPTOMS xxii (1997). 



 

39.2 COLUMBIA JOURNAL OF GENDER AND LAW                         103 

 

 

 

1. Discounting Pain 

 

Accurate and prompt diagnosis usually requires believing a patient’s report of their 

pain, but the trust gap suggests that women are often not afforded that luxury.66 Across 

the board, this discounting of women’s pain leads to more missed and delayed diagnoses. 

For instance, women experience longer delays in the emergency room before being 

treated for acute abdominal pain.67 Women with endometriosis experience, on average, a 

ten-year delay between the onset of symptoms and receiving a diagnosis; because a 

surgical procedure is required to diagnose endometriosis, a physician must first be 

convinced that their patient’s pain is more than just “bad cramps” to send such a patient 

for diagnostic surgery.68 Moreover, about 61% of women who eventually get an 

endometriosis diagnosis are initially told by their healthcare providers that there is 

nothing wrong with them.69  

 
66 See MARTIN PALL, EXPLAINING UNEXPLAINED ILLNESSES 190–91 (“Many psychogenic advocates have 

explained these multisystem illnesses [such as multiple chemical sensitivity and fibromyalgia] as being due to 

some kind of aberrant belief system . . . Typically, the illness belief is claimed to be due not only to the 

sufferer himself but also to . . . medical personnel who ‘validate’ the belief by diagnosing the illness.”); see 

also Alison W. Rebman, Mark J Soloski & John N. Aucott, Sex and Gender Impact Lyme Disease 

Immunopathology, Diagnosis and Treatment, in SEX AND GENDER DIFFERENCES IN INFECTION AND 

TREATMENTS FOR INFECTIOUS DISEASES 337 (Sabra L. Klein & Craig W. Roberts eds., 2015); Reva C. 

Lawrence et al., Estimates of the Prevalence of Arthritis and Other Rheumatic Conditions in the United 

States, Part II, 58 ARTHRITIS & RHEUMATISM 26 (2008); Diane E. Hoffmann & Anita J. Tarzian, The Girl 

Who Cried Pain: A Bias Against Women in the Treatment of Pain, 29 J.L. MED. & ETHICS 13, 20 (2001) 

(“The subjective nature of pain requires health care providers to view the patient as a credible reporter . . . 

[and women are] more likely to have their pain reports discounted as ‘emotional’ or ‘psychogenic’ and, 

therefore, ‘not real.’”). 

 
67 See Ester H. Chen et al., Gender Disparity in Analgesic Treatment of Emergency Department Patients with 

Acute Abdominal Pain, 15 ACAD. EMERGENCY MED. 414, 416 (2008). 

 
68 See Ninet Sinaii et al., High Rates of Autoimmune and Endocrine Disorders, Fibromyalgia, Chronic 

Fatigue Syndrome and Atopic Diseases Among Women with Endometriosis: A Survey Analysis, 17 HUM. 

REPROD. 2715, 2720 (2002); Ruth Hadfield et al., Delay in the Diagnosis of Endometriosis: A Survey of 

Women from the USA and the UK, 11 HUM. REPROD. 878, 879 (1996) (“The considerable delay from the time 

of the initial occurrence of pain symptoms to the diagnosis of endometriosis was reported by women from 

both the USA and the UK, but the mean delay in diagnosis was significantly greater for the American 

women.”). 

 
69 See Rebecca Greene et al., Diagnostic Experience Among 4,334 Women Reporting Surgically Diagnosed 

Endometriosis, 91 FERTILITY & STERILITY 32, 35; see also M. S. Arruda et al., Time Elapsed from Onset of 

Symptoms to Diagnosis of Endometriosis in a Cohort Study of Brazilian Women, 18 HUM. REPROD. 756, 758 

(2003) (“The overall long delay in the present study to reach a diagnosis of endometriosis corroborates with a 

previous observation: ‘Women frequently complain that the diagnosis was made only after months and 



104 COLUMBIA JOURNAL OF GENDER AND LAW 39.2  

But perhaps the most unfortunate example of discounting women’s pain is the 

diagnosis—or lack thereof—of early stage ovarian cancer. It was, until recently, widely 

believed that ovarian cancer has no symptoms until its later stages, when the five-year 

survival rate drops from 92% to 30%.70 The truth, however, is that there are symptoms: 

bloating, pelvic or abdominal pain, difficulty eating, feeling full quickly, and urinary 

urgency or frequency.71 Women with early stage ovarian cancer often have these 

symptoms prior to diagnosis,72 but are frequently misdiagnosed with irritable bowel 

syndrome, urinary tract infections, or simply told that their symptoms are normal for 

menopausal women.73 These symptoms were not recognized by the American Cancer 

Society, Gynecologic Cancer Foundation, or Society of Gynecologic Oncologists until 

2007, when reliable data on the incidence of symptoms was published.74  

 

Additionally, women who report symptoms of chronic pain conditions, which are the 

leading cause of long-term disability in the United States75 and affect more women than 

men,76 are likely to be told by their healthcare providers that their symptoms are “all in 

 
months of being fobbed off or ignored by doctors.’”) (citing Stephen H. Kennedy, What Is Important to the 

Patient with Endometriosis?, 72 BRIT. J. CLINICAL PRAC. SUPPLEMENT 8 (1991)). 

 
70 See OVARIAN, FALLOPIAN TUBE, AND PERITONEAL CANCER: STATISTICS, www.cancer.net/cancer-types/ 

ovarian-fallopian-tube-and-peritoneal-cancer/statistics [https://perma.cc/PD4U-C262]. 

 
71 See Soc’y Gynecologic Nurse Oncologists, CANCER AWARENESS MONTH, https://www.sgno.org/outreach/ 

cancer-awareness-month [https://perma.cc/8KDS-EGWE]. 

 
72 See Lloyd H. Smith, Early Clinical Detection of Ovarian Cancer: A Review of the Evidence, 6 EXPERT 

REV. ANTICANCER THERAPY 1045, 1045 (2014) (finding that a third to half of patients with ovarian cancer 

report early stage symptoms). 

 
73 See DIAGNOSING OVARIAN CANCER, https://www.targetovariancancer.org.uk/health-professionals/gps/ 

diagnosing-ovarian-cancer [https://perma.cc/EF8P-JLE4]. 

 
74 See generally Soc’y Gynecologic Nurse Oncologists, supra note 71; IMAGINIS, CONSENSUS STATEMENT 

RELEASED ON OVARIAN CANCER SYMPTOMS, https://www.imaginis.com/ovarian-cancer-news/consensus-

statement-released-on-ovarian-cancer-symptoms [https://perma.cc/A9W3-8X9K]; Barbara Goff et al., 

Development of an Ovarian Cancer Symptom Index: Possibilities for Earlier Detection, 109 CANCER 221 

(2007). 

 
75 See INST. MED., RELIEVING PAIN IN AMERICA: A BLUEPRINT FOR TRANSFORMING PREVENTION, CARE, 

EDUCATION, AND RESEARCH 5, 89–90 (2011). 

 
76 See Jeffrey S. Mogil, Sex Differences in Pain and Pain Inhibition: Multiple Explanations of a 

Controversial Phenomenon, 13 NATURE REVS. NEUROSCIENCE 859, 859–61 (2012); INST. MED., supra note 

75, at 89–90; MARY LOU BALLWEG ET AL., CAMPAIGN TO END CHRONIC PAIN IN WOMEN, CHRONIC PAIN IN 

WOMEN: NEGLECT, DISMISSAL AND DISCRIMINATION 4 (2010). 



 

39.2 COLUMBIA JOURNAL OF GENDER AND LAW                         105 

 

 

 

their heads,” that they should learn to live with their pain, or that because they “look 

good,” they must not be in as much pain as they allege.77 In fact, the gender bias in 

trusting a patient’s report of pain and symptoms is so ingrained in the medical community 

that female patients report making great efforts to balance assertiveness, emotionality, 

and appearances so that their pain will be taken seriously by practitioners.78 This bias can 

severely affect women’s quality of care and their likelihood of receiving a correct 

diagnosis. 

 

2. Diagnostic Delays 

 

Diagnostic delays can have a far-reaching impact, both emotionally and 

economically. Delays in diagnosing conditions causing disability may result in a patient 

being unable to claim federal protections to which they may otherwise be entitled under 

the Americans with Disabilities Act, which requires accommodations and equal treatment 

in the workplace,79 and Social Security Disability Insurance.80 For some diseases, 

delaying a diagnosis can allow tissue damage to worsen or become permanent, impact the 

odds of remission, and impact how expensive and intensive a patient’s care will be if and 

when they are finally diagnosed.81 Delays in diagnosis accompanied by messaging from 

healthcare professionals that the patient’s symptoms are “all in their head” can also cause 

significant emotional and psychological distress for the patient. 

 

The impact of diagnostic delays falls mainly on female patients, who experience 

longer delays than their male counterparts for nearly every illness.82 These delays are 

 
77 See Pat Anson, Women in Pain Report Significant Gender Bias, NAT’L PAIN REPORTS (Sept. 12, 2014), 

http://nationalpainreport.com/women-in-pain-report-significant-gender-bias-8824696.html [https://perma. 

cc/QVM3-CA4S]. 

 
78 See generally Anne Werner & Kirsti Malterud, It Is Hard Work Behaving as a Credible Patient: 

Encounters Between Women with Chronic Pain and Their Doctors, 57 SOC. SCI. & MED. 1409 (2003); 

Thomas Hadjistavropoulos, Bruce McMurty & Kenneth D. Craig, Beautiful Faces in Pain: Biases and 

Accuracy in the Perception of Pain, 11 PSYCH. & HEALTH 411 (1996). 

 
79 42 U.S.C. § 12111 (2018). 

 
80 42 U.S.C. § 423 (2018). 

 
81 See generally Kristina Forslind, Ingiäld Hafstrom, & Monica Ahlmén, Sex: A Major Predictor of 

Remission in Early Rheumatoid Arthritis?, 66 ANNALS RHEUMATIC DISEASES 46 (2007). 

 
82 See BRAIN TUMOR CHARITY, FINDING MYSELF IN YOUR HANDS: THE REALITY OF BRAIN TUMOR TREATMENT 

AND CARE 6 (2016) (“[O]verall men who participated are more likely to have received a diagnosis in a shorter 

time period and with fewer visits to a doctor than women . . . suggest[ing] a worrying disparity between men 



106 COLUMBIA JOURNAL OF GENDER AND LAW 39.2  

exacerbated by psychosomatic misdiagnoses, which women receive more often than their 

male counterparts,83 and by a discounting of women’s pain and other symptoms. For 

early detection of many diseases, a physician must first be concerned enough about the 

patient’s symptoms to run the appropriate tests. Women experience longer delays in 

testing for many kinds of diseases, including brain, bladder, renal, and other cancers.84 

When there is a psychosomatic misdiagnosis in the way, testing and diagnosis can take 

even longer—up to fourteen times longer, depending on the disease.85 

 

3. Frequent Misdiagnoses 

 

Even when women patients do receive appropriate testing and physical examination, 

they receive less intense screening and treatment than their male counterparts,86 despite 

being equally vulnerable to the top three causes of death—heart disease, stroke, and 

cancer—as men.87 The knowledge gap means that healthcare professionals may make 

assumptions about what diseases are more likely in women and about how certain 

diseases present symptomatically based on their knowledge about the “male model” of 

medicine. The trust gap means healthcare professionals may trust their ideas of the “male 

model” more than they trust their women patients’ reports of their own symptoms. For 

instance, women have different symptoms and risk factors for heart attacks than men and 

 
and women in the speed and ease of obtaining a diagnosis.”); see generally Benjamin Bleicken et al., Delayed 

Diagnosis of Adrenal Insufficiency Is Common: A Cross-Sectional Study in 216 Patients, 339 AM. J. MED. 

SCI. 525 (2010); ANNA KOLE & FRANCOIS FAURISSON, THE VOICE OF 12,000 PATIENTS: EXPERIENCES AND 

EXPECTATIONS OF RARE DISEASE PATIENTS ON DIAGNOSIS AND CARE IN EUROPE (2009); E. Purinszky & O. 

Palm, Women with Early Rheumatoid Arthritis are Referred Later Than Men, 64 ANNALS RHEUMATIC 

DISEASES 1227 (2005). 

 
83 See Netta Levin, Michal Mor & Tamir Ben-Hur, Patterns of Misdiagnosis of Multiple Sclerosis, 5 ISRAELI 

MED. ASS’N J. 489, 490 (2003). 

 
84 See BRAIN TUMOR CHARITY, supra note 82; Nafees U. Din, Age and Gender Variations in Cancer 

Diagnostic Intervals in 15 Cancers: Analysis of Data from the UK Clinical Practice Research Datalink, 10 

PLOS ONE e0127717, 7 (2015); Georgios Lyratzopoulos et al., Gender Inequalities in the Promptness of 

Diagnosis of Bladder and Renal Cancer After Symptomatic Presentation: Evidence from Secondary Analysis 

of an English Primary Care Audit Survey, 3 BRIT. MED. J. OPEN e002861, 6 (2013). 

 
85 See KOLE & FAURISSON, supra note 82, at 48. 

 
86 See Markus Bönte et al., Women and Men with Coronary Heart Disease in Three Countries: Are They 

Treated Differently?, 18 WOMEN’S HEALTH ISSUES 191, 195 (2008). 

 
87 See Wenger, supra note 43, at 558; Luke K. Kim et al., Sex-Based Disparities in Incidence, Treatment, and 

Outcomes of Cardiac Arrest in the United States, 2003-2012, 5 J. AM. HEART ASS’N e003704, 1 (2016). 



 

39.2 COLUMBIA JOURNAL OF GENDER AND LAW                         107 

 

 

 

the typical test for heart attacks is less accurate for women than men.88 Due to the 

knowledge gap, physicians are more likely to miss heart attacks in women because they 

tend to present differently and don’t always correspond to the “male model” of the heart 

attack.89 However, even among patients who present with the typical “male model” heart 

attack symptoms, there is a higher mortality rate among women than among men,90 

potentially because women’s pain is taken less seriously than men’s and their typical 

heart attack symptoms, like chest pain, may be dismissed as anxiety. This suggests that 

the trust gap plays a significant role independent from the knowledge gap.  

 

The outcome of the interaction of the knowledge and trust gaps is that women are 

more likely to be misdiagnosed for illnesses stereotyped as “men’s diseases,” such as 

stroke,91 chronic obstructive pulmonary disorder,92 and heart attacks,93 even though these 

stereotypes are often inaccurate.94 Conversely, women patients are likely to be under-

diagnosed for illnesses which share symptoms with any aspect of the reproductive 

system, whether it be menstruation, pregnancy, motherhood, or menopause. Many 

illnesses are likely to be misdiagnosed if they share any symptoms with menopause, like 

 
88 See Laxmi Mehta et al., Acute Myocardial Infarction in Women: A Scientific Statement From the American 

Heart Association, 133 CIRCULATION 916, 922–23 (2016); Anoop S. V. Shah et al., High Sensitivity Cardiac 

Troponin and the Under-Diagnosis of Myocardial Infarction in Women: Prospective Cohort Study, 350 BRIT. 

MED. J. g7873, 1 (2015); Lori Mosca et al., Fifteen-Year Trends in Awareness of Heart Disease in Women: 

Results of a 2012 American Heart Association National Survey, 127 CIRCULATION 1254, 1258 (2013). 

 
89 See John G. Canto, Association of Age and Sex with Myocardial Infarction Symptom Presentation and In-

Hospital Mortality, 307 JAMA 813, 817 (2012) (finding that men, and especially men who presented with 

chest pain, were significantly more likely to be treated for heart attacks than women, who more often 

presented without chest pain). 

 
90 See Judith H. Lichtman et al., Symptom Recognition and Healthcare Experiences of Young Women with 

Acute Myocardial Infarction, 8 CIRCULATION: CARDIOVASCULAR QUALITY & OUTCOMES S31, S35 (2016). 

 
91 See Newman-Toker et al., supra note 3, at 166 (finding that women’s strokes are more frequently 

misdiagnosed than men’s). 

 
92 See generally Kenneth R. Chapman, Donald P. Tashkin, & David J. Pye, Gender Bias in the Diagnosis of 

COPD, 119 CHEST J. 1691 (2001). 

 
93 See J. Hector Pope et al., Missed Diagnoses of Acute Cardiac Ischemia in the Emergency Department, 342 

NEW ENG. J. MED. 1163, 1168 (2000). 

 
94 See MeiLan K. Han et al., Gender and Chronic Obstructive Pulmonary Disease, 176 AM. J. RESPIRATORY 

& CRITICAL CARE MED. 1179, 1179 (2007) (stating that more women than men die of COPD each year). 



108 COLUMBIA JOURNAL OF GENDER AND LAW 39.2  

hot flashes in non-Hodgkin lymphoma,95 or present with abdominal pain misconstrued as 

menopause or menstrual cramps, like uterine, colon, and bladder cancers.96 The upshot is 

that women experience wrong, delayed, and missed diagnoses for all kinds of diseases, 

across the board. 

 

4. Intersectionality and Impact 

 

It is important to note that gender interacts with other factors like race, 

socioeconomic status, age, and weight, among others. For instance, that healthcare 

professionals tend to underestimate black patients’ pain is widely documented.97 Women 

who are able to get diagnoses for illnesses such as chronic pain conditions, which are 

difficult to diagnose and which present with symptoms that healthcare professionals are 

especially likely to dismiss as trivial or psychosomatic, tend to be white and of higher 

socioeconomic status, having the resources to find a specialist who can diagnose their 

condition.98 Many treatable conditions like Alzheimer’s, which is more prevalent in 

women, are often diagnosed late because symptoms are too casually dismissed as 

‘normal’ signs of aging.99 Women are also more likely than men to be wrongfully told 

their symptoms are a result of being overweight; women are more likely to be advised to 

lose weight by their physicians100 and are advised to do so at smaller amounts of 

 
95 See Angela Epstein, Do You Suffer from Night Sweats? Don’t Blame the Menopause Just Yet: After Being 

Misdiagnosed by Doctors for Two Years, Wendy Discovered She Had Cancer, DAILY MAIL (Dec. 3, 2016), 

www.dailymail.co.uk/health/article-4002952/Do-suffer-night-sweats-Don-t-blame-menopause-just-

misdiagnosed-doctors-two-years-Wendy-discovered-cancer.html [https://perma.cc/2YN7-7Z63]. 

 
96 See DUSENBERY, DOING HARM, supra note 4, at 248; see also Julissa Catalan (as told to), Doctors Told Me 

My Uterine Cancer Was Menopause, PREVENTION (June 7, 2016), www.prevention.com/health/doctors-told-

me-my-uterine-cancer-was-menopause [https://perma.cc/9HSJ-B58G]. 

 
97 See Kelly M. Hoffman et al., Racial Bias in Pain Assessment and Treatment Recommendations, and False 

Beliefs About Biological Differences Between Blacks and Whites, 113 PROCEEDINGS NAT’L ACAD. SCI. 4296, 

4300 (2016); Salimah H. Meghani, Eeeseung Byun, & Rollin Gallagher, Time to Take Stock: A Meta-

Analysis and Systematic Review of Analgesic Treatment Disparities for Pain in the United States, 13 PAIN 

MED. 150, 159 (2012). 

 
98 See Sandra H. Berry et al., Prevalence of Symptoms of Bladder Pain Syndrome/Interstitial Cystitis Among 

Adult Females in the United States, 186 J. UROLOGY 540, 544 (2011). 

 
99 See Karin M. Ouchida & Mark S. Lachs, Not for Doctors Only: Ageism in Healthcare, AGEBLOG (2015), 

www.asaging.org/blog/not-doctors-only-ageism-healthcare [https://perma.cc/ACK5-SK3M]. 

 
100 See Caitlin Anderson et al., Weight Loss and Gender: An Examination of Physician Attitudes, 9 OBESITY 

257, 260–61 (2001); Christopher M. Sciamanna et al., Who Reports Receiving Advice to Lose Weight? 

Results from a Multistate Survey, 160 ARCHIVES INTERN. MED. 2334, 2338 (2000). 



 

39.2 COLUMBIA JOURNAL OF GENDER AND LAW                         109 

 

 

 

‘overweight’ than men101 even though overweight women are more likely than 

overweight men to be metabolically healthy.102 With any of these intersectional identities, 

as well as other identities not discussed here, “doctors too often can’t see past some 

aspect of the patient’s identity that’s considered somehow inherently abnormal,” so their 

symptoms are likely to be misattributed to that ‘abnormal’ identity.103 

 

II. Medical Malpractice Law as a Recourse 

 

Medical malpractice law has been established as an avenue for patients who 

experience medical error to seek justice. This is an important option for patients who 

have been seriously impacted by medical error. Diagnostic error is the most common type 

of malpractice allegation104 and the Institute of Medicine estimates the annual cost of 

medical errors in hospitals to be between $17 and $29 billion.105 The costs of medical 

malpractice suits, on the other hand, while they affect the cost of doctors’ and hospitals’ 

malpractice insurance premiums, do not significantly raise healthcare costs.106 This 

suggests that medical malpractice lawsuits are an economically efficient way for 

individual plaintiffs to distribute the costs of medical error; the economic impact of such 

an error on one plaintiff may be catastrophic,107 but they are not so burdensome when 

 
101 See DUSENBERY, DOING HARM, supra note 4, at 244. On average, women are advised to lose weight when 

they are as little as thirteen pounds overweight while men are not advised to lose weight until they are sixty-

eight pounds overweight. Id.; see also Rebecca M. Puhl, Tatiana Andreyeva & Kelly D. Brownell, 

Perceptions of Weight Discrimination in America, 32 INT’L J. OBESITY 992, 998 (2008) (finding that women 

experience more weight discrimination than men). 

 
102 See Catherine M. Phillips et al., Defining Metabolically Healthy Obesity: Role of Dietary and Lifestyle 

Factors, 8 PLOS ONE e76188, 6 (2013). 

 
103 See DUSENBERY, DOING HARM, supra note 4, at 243. 

 
104 Diagnostic error accounts for 31.8% of all medical malpractice allegations in the National Practitioner 

Data Bank. Adam C. Schaffer et al., Rates and Characteristics of Paid Malpractice Claims Among US 

Physicians by Specialty, 1992-2014, 177 JAMA INTERN. MED. 710, 710 (2017). 

 
105 This includes disability and health care costs, lost income, lost productivity, and other costs of care. See 

PUBLIC CITIZEN: CONGRESS WATCH, MEDICAL MISDIAGNOSIS: CHALLENGING THE MALPRACTICE CLAIMS OF 

THE DOCTORS’ LOBBY 1 (2003). 

 
106 See FURROW ET AL., supra note 7, at 506. 

 
107 See generally Benjamin D. Sommers, Atul Gawande & Katherine Baicker, Health Insurance Coverage 

and Health—What the Recent Evidence Tells Us, 377 NEW ENG. J. MED. 586 (2017) (explaining the negative 

economic impact healthcare costs can have on individual patients). Medical error adds to already burdensome 

healthcare costs by requiring additional medical visits to correct medical errors. 



110 COLUMBIA JOURNAL OF GENDER AND LAW 39.2  

shifted through the insurance system. Given that medical error is the third leading cause 

of death in the United States,108 it is crucial that we maintain an effective system for 

patients and their families and caretakers to recoup these catastrophic costs. 

 

A. An Overview of Medical Malpractice Law 

 

Each state has its own specific medical malpractice laws, but generally, in order to 

prevail in a medical malpractice case, a plaintiff must establish: (1) that the practitioner 

had a duty to the plaintiff, meaning the practitioner had undertaken a professional service 

to the patient; (2) that the practitioner breached that duty by providing service that was 

below the acceptable standard of care; (3) that the plaintiff experienced injury; and (4) 

that the practitioner’s breach of duty directly caused the plaintiff’s injury.109 The focus of 

this study is on the second factor: that the practitioner violated the standard of care.  

 

The standard of care has evolved over time. At first, American courts used an 

evidentiary proof standard which required a plaintiff to present an expert witness from 

the same locality as the defendant to testify regarding current standard medical practice in 

that locality,110 otherwise known as the locality rule. The locality rule dealt with 

differences among physicians in their education and training, since medical school 

curricula were not yet standardized, and in the resources and technology available to 

them.111 However, the locality rule allowed physicians to set a standard for themselves 

within their own profession, and it allowed physicians in small communities who knew 

one another to avoid testifying against each other and thereby escape liability.112 Then, in 

1832, the Connecticut Supreme Court in Landon v. Humphrey coined the language that 

American courts now universally use to assess malpractice claims: A lack of “ordinary 

 
108 See FURROW ET AL., supra note 7, at 507; see also Martin A. Makary & Michael Daniel, Medical Error—

The Third Leading Cause of Death in the U.S., 353 BRIT. MED. J. 2139 (2016). 

 
109 See Barry A. Lindahl, Elements of Medical Malpractice Claim, in 3 MODERN TORT LAW: LIABILITY & 

LITIGATION § 24:1 (2d ed.). See generally, Ellsworth v. United States, No. CV-16-08150-PCT-DLR, 2018 

WL 1784687 (D. Ariz. Apr. 13, 2018); Watson by Leonard v. West Suburban Medical Center, 103 N.E.3d 

895 (Ill. App. Ct. 2018). 

 
110 See Slater v. Baker & Stapleton (1767) 95 Eng. Rep. 860. 

 
111 See generally STEVEN NEHMER, THE NEW JERSEY MEDICAL MALPRACTICE LIABILITY INSURANCE CRISIS OF 

2002 3, 21 (2005); Allen D. Spiegel & Florence Kavaler, America’s First Medical Malpractice Crisis, 1835-

1865, 22 J. COMMUNITY HEALTH 283 (1997). 

 
112 See Leonard Berlin, Radiologic Malpractice Litigation: A View of the Past, a Gaze at the Present, a 

Glimpse of the Future, 181 AM. J. ROENTGENOLOGY 1481, 1482–83 (2003). 



 

39.2 COLUMBIA JOURNAL OF GENDER AND LAW                         111 

 

 

 

diligence” on the practitioner’s part entitles an injured plaintiff to recover.113 The term 

“ordinary diligence,” without reference to local custom or practice, along with the 

establishment of the American Medical Association in 1847 and the resulting increased 

standardization of the medical field, prompted states to begin shifting their legal 

standards to a national rather than a locality rule.114 

 

At the same time, medical malpractice litigation was becoming more frequent and 

case law slowly helped establish some regulations for a poorly regulated profession.115 As 

medical malpractice litigation flourished and as most states began using a national rather 

than locality rule, practitioner liability expanded and practitioners began practicing 

defensive medicine—ordering more tests and more expensive screening to avoid later 

claims that they did not exercise due care.116 Two recent studies found that between 

73%117 and 93% of private practitioners in the United States admitted to practicing 

defensive medicine for fear of litigation despite the fact that medical error must be 

negligent in order for a plaintiff to recover at law,118 as measured against what “an 

ordinarily prudent physician would do under the same or similar circumstances.”119 

 

Now, in most states, violating the standard of care means a practitioner’s medical 

service was substandard as measured against a national “accepted medical practice” 

 

 
113 9 Conn. 209 (1832), overruled in part by Crosby v. Fitch, 12 Conn. 410 (1838). 

 
114 Id. 

 
115 See Robert I. Field, The Malpractice Crisis Turns 175: What Lessons Does History Hold for Reform?, 4 

DREXEL U.L. REV. 7, 11–12 (2011). 

 
116 See Sonal Sekhar Miraj & Navya Vyas, Defensive Medicine: A Bane to Healthcare, 3 ANNALS MED. & 

HEALTH SCI. RES. 295, 295 (2013). 

 
117 Id. 

 
118 JACKSON HEALTHCARE, PHYSICIAN STUDY: QUANTIFYING THE COST OF DEFENSIVE MEDICINE, https:// 

jacksonhealthcare.com/media-room/surveys/defensive-medicine-study-2010/ [https://perma.cc/5S2L-AYSE] 

(“Fear of litigation has been cited as the driving force behind defensive medicine. Defensive medicine is 

especially common in the United States of America, with rates as high as 79% to 93%, particularly in 

emergency medicine, obstetrics, and other high-risk specialties.”). 

 
119 See Apodaca v. Miller, 281 S.W.3d 123 (Tex. Ct. App. 2008); Moore v. Sutherland, 107 S.W.3d 786 (Tex. 

Ct. App. 2003). 



112 COLUMBIA JOURNAL OF GENDER AND LAW 39.2  

benchmark.120 While plaintiffs are still required to provide an expert to testify as to the 

standard of care in a particular case, and the defendant practitioner may also offer an 

expert witness,121 plaintiffs can use an expert of the same specialty as the defendant from 

anywhere in the country.122 However, a minority of states still use the locality rule, where 

experts need only testify to what a practitioner in that community should have done in 

similar circumstances.123 Moreover, even in states with a national rule, medical 

malpractice standards are largely protective of physicians. Practitioners need only be 

“minimally competent,” and a plaintiff’s expert must be able to testify that the 

defendant’s actions fell below that minimal standard, not just below an optimal standard 

of care.124 

 

B. The Gendered Impact of Medical Malpractice Law 

 

As it currently stands, the legal standard for medical malpractice may not equally 

protect male and female patients. As stated in the seminal case McCourt v. Abernathy, 

“generally qualified physicians may differ as to what constitutes a preferable course of 

treatment. Such differences due to preference . . . do not amount to malpractice.”125 

Because knowledge in the medical community about “female-pattern” symptom 

presentations for “male diseases”126 and about illnesses that predominantly affect women 

is spotty at best,127 and because the trust gap is so pervasive, the “minimally competent” 

practitioner may not be aware of recent developments in knowledge about women’s 

health and may not give much weight to women’s reports of their symptoms. Listening to 

 
120 See Faicco v. Golub, 938 N.Y.S.2d 105, 106 (App. Div. 2012). 

   
121 See PETER P. BUDETTI & TERESA M. WATERS, MEDICAL MALPRACTICE LAW IN THE UNITED STATES 2–3 

(Henry J. Kaiser Fam. Found. ed., 2005). 

 
122 See Hall v. Hilbun, 466 So. 2d 856, 866 (Miss. 1985). 

 
123 See BUDETTI & WATERS, supra note 121, at 6. 

 
124 See Peter Moffett & Gregory Moore, The Standard of Care: Legal History and Definitions: The Bad and 

Good News, 12 W.J. EMERGENCY MED. 109, 112 (2011). 

 
125 McCourt ex rel. McCourt v. Abernathy, 318 S.C. 301, 307 (1995). 

 
126 See TEDxWomen, The Single Biggest Health Threat Women Face, TED (Dec. 2011), 

https://www.ted.com/talks/noel_bairey_merz_the_single_biggest_health_threat_women_face?language=en 

[https://perma.cc/SGF2-JJU5]. 

 
127 See Stephen Silberstein, quoted in Joanna Kempner, NOT TONIGHT: MIGRAINE AND THE POLITICS OF 

GENDER AND HEALTH 12 (2014); Pamela Weintraub, CURE UNKNOWN: INSIDE THE LYME EPIDEMIC 66 (2013). 



 

39.2 COLUMBIA JOURNAL OF GENDER AND LAW                         113 

 

 

 

and trusting women patients should be part of a minimal standard of care, but in the 

current state of medical malpractice law, such distrust likely is not a deviation from the 

basic standard of care as long as a practitioner takes what most “ordinary” practitioners 

would deem to be reasonable action in response to their women patients. As such, women 

patients alleging missed, wrong, or delayed diagnoses may have more difficulty than their 

male counterparts in bringing and prevailing in medical malpractice suits. 

 

However, the scope of legal protections for patients against gender discrimination in 

the medical setting may change in the near future. The Patient Protection and Affordable 

Care Act (ACA), passed in 2010,128 included a provision129 prohibiting discrimination in 

health programs based on four existing civil rights laws, including Title IX of the 

Education Amendments of 1972.130 In 2016, the Department of Health and Human 

Services (HHS) issued a rule interpreting this provision as prohibiting discrimination in 

healthcare on the basis of sex stereotypes, gender identity, and pregnancy-related 

conditions.131 A few months later, though, Judge O’Connor of the Northern District of 

Texas issued a nation-wide preliminary injunction on the HHS rule in Franciscan 

Alliance v. Burwell, holding that the HHS rule’s inclusive definition of “sex” likely 

exceeded HHS’s authority as granted by Congress in the ACA and that the rule likely 

violated plaintiffs’ religious liberties as protected by the Religious Freedom Restoration 

Act.132 Following this ruling, the Attorney General withdrew the Department of Justice’s 

guidance regarding the use of the word “sex” in Title IX and its implications for 

discrimination, which restricted the authority of Title IX as cited by the HHS in its earlier 

2016 rule.133 This renders the expanded gender protections of the ACA and the 2016 

HHS rule essentially toothless as applied to the healthcare context, but future litigation or 

changes in the executive branch may further the ACA’s goal of protection against 

discrimination. 

 

 

 
128 See 42 U.S.C. § 18001 et seq. (2018). 

 
129 See 42 U.S.C. § 18116 (2018). 

 
130 See 20 U.S.C. § 1681 et seq. (2018). 

 
131 See 81 Fed. Reg. 31,375 (May 18, 2016). 

 
132 227 F. Supp. 3d 660, 695 (N.D. Tex. 2016). 

 
133 See Press Release, Dep’t of Justice, Press Release 17-214, Statement by Attorney General Jeff Sessions on 

the Withdrawal of Title IX Guidance (Feb. 22, 2017) (on file with the U.S. Department of Justice). 



114 COLUMBIA JOURNAL OF GENDER AND LAW 39.2  

For now, the legal ramifications of the trust gap are still very much in effect. Take, 

for instance, the case of Wilburn v. Cleveland Psychiatric Institute.134 When Edna 

Wilburn experienced her first stroke, her attending practitioners did not complete the 

standard stroke protocol despite Edna’s exhibition of standard stroke symptoms.135 

Instead, a hospital psychiatrist determined that Edna must be “suffering from conversion 

reaction disorder, a mental illness causing physical impairment without any organic 

basis.”136 After Edna’s second stroke and return to the hospital, she was again diagnosed 

with conversion disorder and referred for inpatient psychiatric care, although this time 

her physician indicated further neurological testing was necessary.137 But Edna’s 

inpatient psychiatric institute did not follow through on this testing despite a worsening 

of symptoms. Edna was misdiagnosed with severe depression and later released for 

outpatient care.138 After her release, Edna suffered a third severe and disabling stroke, 

after which she was finally diagnosed with a rare blood disorder that causes strokes.139 

Edna won her case at the trial court, but the court of appeals reversed, holding that the 

appropriate issue to consider was whether the defendants violated the standard of care in 

failing to rule out conversion disorder rather than failing to diagnose Edna’s blood 

disorder.140 The implication of the court was that even if the practitioners violated the 

standard of care by not ruling out conversion disorder, they may not have correctly 

diagnosed Edna’s blood disorder and prevented subsequent strokes, and thus the 

practitioners’ error would not have proximately caused Edna’s damages. The court 

seemingly ignored the fact that the practitioners’ original diagnosis of conversion order 

stigmatized Edna and delayed testing and treatment that could have, at minimum, treated 

the worsening symptoms of her strokes, if not their underlying cause.  

 

Unfortunately, Edna is not alone. In Pleasants v. Alliance Corp., a teenage girl who 

presented in the emergency room with severe abdominal pain was released within two 

 

 
134 743 N.E.2d 515 (Ohio Ct. App. 2000). 

 
135 Id. at 516. 

 
136 Id. 

 
137 Id. 

 
138 Id. 

 
139 Id. 

 
140 Wilburn, 743 N.E.2d at 519. 



 

39.2 COLUMBIA JOURNAL OF GENDER AND LAW                         115 

 

 

 

hours to die at home from cardiac arrest induced by severe stomach infection.141 Her 

attending physician, underestimating her pain, assumed run-of-the-mill gastroenteritis, a 

swelling of the lining of the stomach and intestines.142 The infection that killed the 

decedent was rare, as was Edna’s blood disorder,143 but the plaintiffs in Pleasants did not 

allege that the defendant’s deviation from the standard of care was in failing to diagnose 

this infection right away; rather, plaintiffs alleged that the violation of the standard of 

care was in discharging the decedent rather than taking her symptoms seriously and 

keeping her for observation.144 Nonetheless, plaintiffs did not prevail on this issue, as the 

court held that the jury reasonably concluded the physician took appropriate actions 

based on the “multiple methods of treatment” available for the symptoms with which the 

patient presented.145 Similarly, in Evers v. Dollinger, the trial court granted summary 

judgment for a physician who told his female patient to “stop worrying and go home and 

relax,” failing to diagnose her breast cancer before it metastasized.146 While Evers was 

overturned on appeal,147 the majority of patients and plaintiffs do not have the time and 

resources to pursue an appeal. 

 

C. Uncertainty of Assessing Gender Bias in Medical Malpractice Cases 

 

Cases like Wilburn, Pleasants, and Evers are present across states and over time, but 

it should be noted that it is difficult to assess how commonplace such analyses as that in 

Edna’s case are. There are often other dispositive issues on which the court makes its 

decision and therefore the impact of gender bias on the final outcome is variable, 

depending on the number and complexity of issues of law present in a given case. For 

instance, in Russo v. Phoenix Internal Medicine Associates, PC, the defendants failed to 

diagnose the decedent with a heart infection, instead minimizing and misattributing her 

symptoms to complications due to asthma.148 The ruling in Russo, though, hinged on the 

 
141 209 W. Va. 39, 41 (2000). 

 
142 Id. 

 
143 Wilburn, 743 N.E.2d at 516. 

 
144 Pleasants, 209 W. Va. at 42. 

 
145 Id. at 47–48. 

 
146 95 N.J. 399, 402 (1984). 

 
147 Id. 

 
148 950 A.2d 559, 561 (Conn. App. Ct. 2008). 



116 COLUMBIA JOURNAL OF GENDER AND LAW 39.2  

validity of the plaintiff’s expert testimony, so the court never reached the issue of 

defendants’ violation of the standard of care.149 It is unclear how the issue of deviation 

from the standard of care would have been assessed by the court. 

 

It is also possible that cases which reach a jury—a small minority of medical 

malpractice cases150—may be more likely to achieve justice for women patients. A 

balanced jury, women among them, may be more trusting of or more sympathetic 

towards women patients who have been unjustly discounted or distrusted in ways that led 

to their medical injuries. However, the available data on whether judges and juries differ 

in the damage awards they dole out to plaintiffs is contradictory.151 Moreover, jury 

verdicts in medical malpractice cases typically receive significant cuts after trial, meaning 

plaintiffs receive less than they were originally awarded.152 

 

Thus, given the high degree of variability in medical malpractice cases, an aggregate 

analysis of case outcomes by gender is necessary to shed light on the potential large-scale 

impact of the knowledge and trust gaps on legal avenues for patient recourse. It is 

important to note, though, that there is an intermediate step in which the knowledge and 

trust gaps may also come into play. After a patient experiences malpractice, their 

likelihood of obtaining adequate representation depends on the value of their case. There 

is a significant subset of patients who experience medical malpractice but who have 

difficulty bringing suit because the economic value of their case may not be sufficient to 

attract the attention of a lawyer due to the patient’s age, as damages are often calculated 

in terms of lost future wages or productivity, or due to the lesser economic impact of the 

 

 
149 Id. at 566. 

 
150 Only about 4% of malpractice cases filed in federal district courts reach a verdict; nearly a quarter are 

settled. Medical Malpractice Litigation Analytics, WESTLAW EDGE, https://1.next.westlaw.com/ (follow 

“Litigation Analytics” link; then search for “medical malpractice,” then select “medical” hyperlink under 

types of cases). 

 
151 See Theodore Eisenberg & Michael Heise, Judge‐Jury Difference in Punitive Damages Awards: Who 

Listens to the Supreme Court?, 8 J. EMPIRICAL LEGAL STUD. 325, 325 (2011) (finding significant differences 

between the damages amounts awarded by judges and juries). Cf. Theodore Eisenberg et al., Juries, Judges, 

and Punitive Damages: An Empirical Study, 87 CORNELL L. REV. 743, 779 (2002) (finding no significant 

difference between damages awarded by judges and juries). 

 
152 See David A. Hyman et al., Do Defendants Pay What Juries Award? Post‐Verdict Haircuts in Texas 

Medical Malpractice Cases, 1988–2003, 4 J. EMPIRICAL LEGAL STUD. 3, 3 (2007). 



 

39.2 COLUMBIA JOURNAL OF GENDER AND LAW                         117 

 

 

 

damage caused by the malpractice.153 Women and the elderly bear the brunt of this 

difficulty in attracting lawyers’ attention.154 That is to say, the impact of the knowledge 

and trust gaps on patients’ access to legal justice is likely larger than what is visible in the 

outcome of cases which are filed and pursued. 

 

III. An Empirical Aggregate Analysis of Gender Bias in Medical Malpractice 

Lawsuits for Delayed, Wrongful, or Missed Diagnoses 

 

The present study was designed to assess whether there is a gendered difference in 

the outcomes of medical malpractice suits alleging wrongful or missed diagnoses, given 

women’s historical underrepresentation in medical studies, the distrust of women patients 

in the medical setting, and the legal standard for medical malpractice suits. As discussed 

in Part II.A., the largest centralized, publicly accessible database of medical malpractice 

lawsuits is the National Practitioner Data Bank (NPDB), which collects data on payments 

made to patients alleging medical malpractice. The present study used NPDB data to 

assess: 

 

1. Whether there is a gendered difference in the average recovery amount for 

medical malpractice plaintiffs alleging delayed, wrong, or missed diagnoses; 

2. Whether such a gender gap, if present, varies by the age of the patient; 

3. Whether such a gender gap, if present, varies by the patient’s health outcome 

as a result of the alleged malpractice; 

4. Whether such a gender gap, if present, varies by the state in which the patient 

accessed their medical care, based on that state’s economic, legal, and public 

health landscape; 

5. And whether such a gender gap, if present, has changed over time, along 

with the accumulation of knowledge about women’s health. 

 

A. Data Collection  

 

The dataset analyzed in this study, hereinafter referred to as the compiled dataset, 

was based on NPDB data and supplemented with data from other sources, all of which 

are accessible online by the general public. NPDB data was used because, according to 

the Code of Federal Regulations, any entity paying a judgment or settlement on behalf of 

 
153 Marshall Allen & Olga Pierce, Ten Patient Stories: When Attorneys Refused My Medical Malpractice 

Case, PROPUBLICA (Jan. 9, 2014), https://www.propublica.org/article/ten-patient-stories-when-attorneys-

refused-my-medical-malpractice-case [https://perma.cc/EZ6M-TF4U]. 

 
154 Id. 



118 COLUMBIA JOURNAL OF GENDER AND LAW 39.2  

a practitioner in a medical malpractice lawsuit must report that payment, along with other 

information, to the NPDB.155 This database is therefore large, centralized, and reliable. 

The compiled dataset contains 47,293 observations, each of which represents a case 

reported to the NPDB for which a payment was made on behalf of a practitioner to a 

plaintiff between 2004 and 2018. Supplemental data was then collected from various 

sources to account for contextual variables that may have an influence on the relationship 

between patient gender and recovery amounts in medical malpractice suits alleging 

missed, delayed, or wrongful diagnoses. Supplemental data was collected from the 

Centers for Disease Control Interactive Atlas of Heart Disease and Stroke Tables, the 

Kaiser Family Foundation’s State Profiles for Women’s Health, the Courtroom Statistics 

Project, Gallup, the U.S. Decennial Census, and the American Community Surveys. 

Supplemental data was then mapped onto the NPDB data, forming the complete 

compiled dataset. 

 

The final addition to the compiled dataset was an index representing how receptive a 

particular state’s court system is to medical malpractice lawsuits. This index was 

constructed based on state statutes. States were given a score on the index based on the 

statute of limitations for medical malpractice suits; whether the state currently has, or had 

at any point between 2004 and 2018, a cap on damages and, if so, whether the cap is for 

noneconomic damages only or for total damages; and whether the state currently uses a 

locality rule or used a locality rule at any point between 2004 and 2018. 

 

STATA software was used to analyze the compiled dataset. Table 1 below presents 

detailed information about the distribution of all variables in the compiled dataset. Note 

that each observation represents an individual plaintiff’s case. Also note that recovery 

amount and state population are logged variables which use the natural log of the amount 

which the plaintiff was paid and the state population, respectively, which limits the effect 

of extreme values on regression estimates. In the table below, the abbreviation “obs.” is 

used to refer to observations, or the number of cases in the NPDB for which a particular 

variable is available, and the abbreviation “std. dev.” is used to refer to standard 

deviation, or the amount of variation in the data set. 

 

 

 

 

 

 

 
155 Reporting Medical Malpractice Payments, 45 C.F.R. § 60.7 (2019). 



 

39.2 COLUMBIA JOURNAL OF GENDER AND LAW                         119 

 

 

 

Table 1. Descriptive Statistics of Sample 

 
Variable Obs. Mean Std. Dev. Min Max 

Case Characteristics 

Patient age 46,032 42.08136 18.43256 1 90  

Patient gender  

Female = 1; Male = 2 

47,293 1.477343 0.4994917 1 2  

Patient health outcome 

Categories of injury scaled from 1, 

least serious, to 4, death 

46,905 3.060761 0.9409799 1 4  

Duration of case, in years 47,293 5.044298 2.777423 0 105  

Total payment to plaintiff 47,293 361,326.9 482,729 50 22,500,000 

Total payment (logged) 47,293 12.12139 1.325691 3.912023 16.92903 

Year of settlement or judgment 47,293 2010.229 4.234526 2004 2018 

Case settled or went to judgment 

Judgment = 1; Settlement = 2 

47,293 1.977502 0.1482982 1 2  

Type of misdiagnosis alleged 

Diagnoses allegations categorized as 

missed, 1, wrongful, 2, or delayed, 3 

47,293 145.1765 64.36294 1 3  

Practitioner Characteristics 

Practitioner age 47,149 44.77804 11.03455 10 80  

Year of professional school graduation 47,293 1979.025 12.38468 1910 2010  

Practitioner field of license 47,293 1.145032 0.4808835 1 4 

State Characteristics 

State 

Numbered in alphabetical order 

47,293 25.88603 13.51809 1 51  

State congeniality index 47,293 8.911721 1.512993 5 13  

Population 28,201 1.37e+07 9,856,671 514,044 38,800,000 

Population (logged) 28,201 16.13034 0.8602984 13.15006 17.4729 

Percent Black 28,201 12.73307 7.453612 0.45 55.5 

Percent Hispanic 28,201 15.2274 11.40427 0.99 47.67 

Percent aged 65+ 28,201 13.49496 2.016766 6.65 19.06 

Percent female aged 65+ 28,201 15.22707 2.120337 7.38 20.54  

Percent urban 28,201 81.56247 11.46776 38.2 100  

Civil caseload per 100k residents 5,547 5,307.5 2,552.08 1,874.834 17,654.39 

Medical malpractice caseload per 100k 

residents 

1,992 106.0112 591.441 1.45895 3,682.4 

Medical malpractice jury trial rate 2,574 0.049033 0.022836 0 0.19587629 

Percent lacking health insurance 3,694 11.85003 3.829458 3 20  

Female cardiac incident death rate 28,201 210.4925 30.25709 40.5 316.9 

Female preventable death rate 28,201 42.43126 9.285987 19.2 86.5 

Gini coefficient 28,201 0.4670634 0.0195129 0.406 0.543 

Per capita income 28,201 31091.86 4142.548 21,494 49,213.18 

Political affiliation (by majority) 

Democratic = 1; Republican = 2 

37,556 1.70974 0.4538883 1 2  

 



120 COLUMBIA JOURNAL OF GENDER AND LAW 39.2  

B. Results 

 

Analysis of the compiled dataset revealed that patient gender impacts recovery 

amounts in diagnosis-related medical malpractice cases. More specifically, patient gender 

interacts with contextual factors—specifically, patient age, patient health outcome, and 

state-specific factors—to influence recovery amounts.  

 

The average recovery amount for female patients in this sample was $358,860.10, 

compared to an average recovery amount for male patients of $364,027.90. This gender 

gap was analyzed using three types of statistical models assessing the impact of 

contextual variables on recovery amounts and the gender gap in recovery amounts: 

practitioner characteristics models, patient characteristics models, and state 

characteristics models. The following subsections discuss each of these statistical models 

and explain the usefulness of each model in predicting recovery amounts in diagnosis-

related malpractice suits. 

 

Given that the payment amounts were highly skewed and there was not a normal 

distribution for this variable, the natural log of this variable, hereinafter referred to as the 

“logpay variable,” was used to generate all models. Each model is based on a linear 

regression.156 All models discussed here use fixed effects for the state in which a case was 

brought and the year in which it was resolved;157 controls for the duration of the case, the 

natural log of the population of the state in which the case arose, and the percentages of 

that state’s population living in urban areas, identified as black, and identified as 

Hispanic; and includes probability weights for the percent of the national population 

living in the state in which the case occurred.158 For all models, standard errors were 

clustered by state.159 Table 2 below presents the results of each linear regression model. 

 
156 A linear regression is a statistical test that calculates how much of the variance in the dependent 

variable—in this case, the logpay variable—is explained by the independent variables, or the factors being 

assessed for their influence on the dependent variable. 

 
157 Fixed effects account for variables that do not change, like the state variable, or that change at a constant 

rate over time, like the year variable. These variables have a consistent impact on the dependent variable—in 

this case, the logpay variable—and the statistical models discussed here account for that consistent impact. 

 
158 This accounts for the increased likelihood of medical malpractice cases in a state with a higher population 

density. In other words, the models discussed here account for the fact that there will inevitably be more 

malpractice cases filed and resolved in New York than in North Dakota simply because there are more people 

in New York than in North Dakota. 

 
159 Standard error is the amount of variation in the probabilities of different possible outcomes in an 

experiment. Standard errors are clustered when some data points are related to each other because a variable 



 

39.2 COLUMBIA JOURNAL OF GENDER AND LAW                         121 

 

 

 

Table 2. Linear Regression Results 

 
Model 1 2 3 4 5 6 7 8 

Pt gender 0.084⁑ 

(0.036) 

0.006 

(0.028) 

-0.201⁑ 

(0.045) 

0.010  

(0.077) 

0.065* 

(0.030) 

0.099⁑ 

(0.015) 

-0.003 

(0.087) 

-0.039 

(0.042) 

Pr age -0.001 

(0.002) 

      -0.005 

(0.003) 

Pr grad yr 0.001 

(0.001) 

      0.006 

(0.005) 

Pr field of license -0.322⁑ 

(0.026) 

      -0.376⁑ 

(0.045) 

Pt allegation -0.001⁑ 

(0.000) 

0.000 

(0.000) 

-0.000 

(0.000) 

0.001 

(0.001) 

-0.000 

(0.000) 

-0.000 

(0.000) 

 -0.000 

(0.000) 

Pr/pt age 

difference 

-0.089⁑ 

(0.010) 

-0.063* 

(0.251) 

-0.009 

(0.059) 

-0.032 

(0.054) 

-0.017 

(0.018) 

-0.033 

(0.038) 

 -0.158 

(0.095) 

Pt age  -0.004⁑ 

(0.001) 

-0.003 

(0.003) 

-0.003 

(0.003) 

-0.011⁑ 

(0.001) 

-0.016⁑ 

(0.003) 

 -0.002 

(0.003) 

Pt outcome  0.451⁑ 

(0.039) 

     0.576⁑ 

(0.044) 

Suit outcome  -0.518⁑ 

(0.107) 

-0.288 

(0.218) 

-0.524⁑ 

(0.174) 

-0.666⁑ 

(0.200) 

-0.489⁑ 

(0.173) 

 -0.162 

(0.372) 

State congeniality 

index 

      6.506⁑ 

(2.104) 

6.019⁑ 

(1.238) 

% uninsured       -0.099 

(0.066) 

-0.061 

(0.032) 

State political 

party majority 

      -37.541* 

(16.491) 

-55.789⁑ 

(10.723) 

% pop over 65       -0.721 

(0.382) 

-0.602⁑ 

(0.219) 

State per capita 

income 

      0.000 

(0.000) 

0.000 

(0.000) 

State gini index       1.102 

(17.188) 

4.365 

(13.025) 

State rate of fatal 

cardiac events in 

females 

      0.003 

(0.012) 

-0.002 

(0.006) 

State rate of 

preventable 

deaths in females 

      0.001 

(0.065) 

-0.002 

(0.040) 

State civil 

caseload per 100k 

residents 

      -0.000 

(0.000) 

-0.000 

(0.000) 

 

 
is identical for a group of data points. For example, in the compiled data set, the state variable is the same for 

many data points—all the cases which were filed in a particular state. 



122 COLUMBIA JOURNAL OF GENDER AND LAW 39.2  

Duration of case 0.040⁑ 

(0.015) 

0.010 

(0.013) 

0.142⁑ 

(0.035) 

0.012 

(0.035) 

-0.039⁑ 

(0.007) 

-0.035⁑ 

(0.010) 

0.345 

(0.182) 

0.191* 

(0.091) 

Population 

(logged) 

-0.369 

(0.789) 

-0.418 

(0.742) 

-2.279 

(2.496) 

1.448 

(1.337) 

1.976 

(1.016) 

0.088 

(1.264) 

6.773 

(5.209) 

15.359⁑ 

(3.234) 

% Urban 0.016 

(0.016) 

0.015 

(0.014) 

-0.029 

(0.054) 

-0.015 

(0.049) 

0.023 

(0.025) 

0.006 

(0.028) 

1.024 

(0.565) 

1.547⁑ 

(0.399) 

% Black 0.131⁑ 

(0.043) 

0.071⁑ 

(0.025) 

0.053 

(0.127) 

-0.007 

(0.097) 

-0.132⁑ 

(0.045) 

0.175 

(0.088) 

1.645* 

(0.744) 

1.038⁑ 

(0.347) 

% Hispanic -0.016 

(0.025) 

-0.026 

(0.020) 

0.038 

(0.110) 

-0.066 

(0.056) 

-0.035 

(0.023) 

-0.047 

(0.039) 

-0.118 

(0.198) 

0.002 

(0.168) 

Constant 11.990 

(8.864) 

15.731 

(9.063) 

41.518 

(31.609) 

-7.437 

(17.227) 

-12.503 

(12.480) 

12.231 

(15.418) 

-168.058 

(91.972) 

-302.121⁑ 

(64.392) 

Observations 28,063 27,596 1,727 4,120 7,872 9,111 2,572 2,516 

R-squared 0.1544 0.2304 0.1939 0.2359 0.1441 0.2106 0.1419 0.3182 

 

Significance: * = p<0.05; ⁑ = p<0.01 

“Pr” = practitioner 

“Pt” = patient 

“Pop” = population 

Note: Robust standard errors are in parentheses. 

 

1. Practitioner Characteristics Models 

 

This model, labeled as Model 1 in Table 2, was designed to assess the impact of 

practitioner characteristics on patients’ recovery amounts. The practitioner characteristics 

accounted for in this model were the practitioner’s age, year of graduation from 

professional school, field of license, and the specific allegation made against the 

practitioner. The practitioner field of license variable included four categories: (1) 

primary and general care—allopathic physicians, physician residents, osteopathic 

physicians, and osteopathic physician residents; (2) specialists—dietitians, podiatrists, 

audiologists, vision and dental care service providers, and respiratory therapists; (3) 

nursing staff and physician assistants; and (4) emergency practitioners. The specific 

allegation variable included three categories: (1) failure to diagnose, (2) delay in 

diagnosis, and (3) wrong diagnosis. Patient gender was a significant predictor of the 

logpay variable, meaning it significantly predicted the amount of payment to plaintiffs, 

with male patients recovering more than their female counterparts. Patients who were 

younger than their practitioners recovered significantly more than their counterparts who 

were older than their practitioners; plaintiffs whose cases took longer recovered 

significantly more than their counterparts whose cases were resolved more quickly; and 

patients in states with higher proportions of black residents recovered significantly more 

than their counterparts. This model explains 15.44% of the variance in the logpay 



 

39.2 COLUMBIA JOURNAL OF GENDER AND LAW                         123 

 

 

 

variable. In other words, the variables in this model explain 15.44% of the differences in 

logpay values, or the payment outcomes for plaintiffs included in the compiled dataset.160 

 

The Model 1 linear regression was then used to calculate a predicted value which is 

graphically represented in Figure 1 below for male and female patients. The predicted 

value is an expression of the logpay variable, which here is being used as a proxy for 

plaintiff recovery,161 based on all variables accounted for in Model 1. Figure 1 shows 

that, while Model 1 shows a statistically significant difference in payment amounts to 

plaintiffs based in part on patient gender, this difference does not appear to be very large, 

indicating that other variables may play a significant role in either amplifying or 

mitigating the role of patient gender in predicting plaintiff recovery. 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 
160 It is important to note two things here: First, while 15% might seem like a low number, given that human 

behavior is difficult to predict and given that this study is looking for percent differences in payment 

amounts—a continuous variable—a model explaining 15% of the variability in payment amounts has 

substantial explanatory power. See, e.g., LINEAR REGRESSION MODELS, https://people.duke.edu/~rnau/ 

rsquared.htm [https://perma.cc/J9LC-SJAW]. Second, in a strictly technical sense, the variability referred to 

here actually refers to the logpay variable, not the payment variable; however, the logpay variable is used as a 

proxy for payment amounts because the payment variable itself has a highly skewed distribution. 

 
161 See the explanation provided supra note 160. 



124 COLUMBIA JOURNAL OF GENDER AND LAW 39.2  

Figure 1. Practitioner Characteristics Model 

 

 
 

2. Patient Characteristics Models 

 

This set of models was designed to assess the impact of patient characteristics on 

patients’ recovery amounts. The patient characteristics measured here were patient age; 

patient health outcome—whether the patient experienced minor injury (purely emotional 

or minor and temporary physical injury), moderate injury (major but temporary physical 

injury or minor permanent injury), major injury (significant permanent injury), or death; 

whether the patient’s case settled or received a judgment; and the type of misdiagnosis 

alleged—missed, delayed, or wrong diagnosis. 

 

In the first patient characteristics model, labeled as Model 2 in Table 2 above, 

patients who did not settle; younger patients; patients who suffered worse health 

outcomes; and patients in states with a higher proportion of black residents recovered 

significantly more than their counterparts. Patient gender, however, was not statistically 

significant, meaning it did not help predict a plaintiff’s recovery amount. This model 

explained 23.04% of the variance in the logpay variable, or 23.04% of the variability in 

the recovery amounts received by plaintiffs included in the compiled dataset. 

 

10.5

10.7

10.9

11.1

11.3

11.5

11.7

11.9

12.1

12.3

12.5

2 0 0 4 2 0 0 6 2 0 0 8 2 0 1 0 2 0 1 2 2 0 1 4 2 0 1 6 2 0 1 8

P
re

d
ic

te
d

 V
al

u
e

Year

Pract i t ioner Character i s t ics  Model

Male Female



 

39.2 COLUMBIA JOURNAL OF GENDER AND LAW                         125 

 

 

 

But a closer look at the data reveals that the lack of a statistically significant gender 

gap in this model is actually the result of a reversed gender gap among patients under 

twenty years old, which masks the effect of the gender gap among adult patients. Female 

minors recovered significantly more than male minors (p=.000), with the average 

recovery amount for female minors being $386,518.10, compared to $362,263.50 for 

male minors. But among patients over twenty years old, the gender gap in recovery 

amounts was as hypothesized based on the literature on the knowledge and trust gaps. 

Among adults under sixty years old, female patients recovered significantly less than 

their male counterparts (p=.001), with the mean recovery amount for a female patient 

being $375,982.10, compared to $384,188 for a male patient. Among patients over sixty 

years old, the gender gap was even more pronounced. Female patients recovered 

significantly less than their male counterparts (p=.000), with female patients recovering 

an average of $218,087 compared to male patients’ $247,802.40. This amounts to an 

average 13.6% difference in recovery amounts based on patient gender. Figure 2 below 

presents a graphical representation of this data, showing that there is a statistically 

significant gender gap in recovery amounts for each aforementioned age group. 

 

Figure 2. Average Recovery Amount by Age Group 

 

 
 

This interaction between patient age and patient gender indicates that gendered 

effects may be masked by other variables which interact with patient gender. Such an 

interaction is evident also in the analysis of patient health outcome: The gender gap in 

0

50000

100000

150000

200000

250000

300000

350000

400000

450000

Minors Adults Under 60 Adults Over 60

U
S

D

Average Recovery Amount by Age Group

Female Male

p=.000 p=.001

p=.000



126 COLUMBIA JOURNAL OF GENDER AND LAW 39.2  

recovery amounts was significantly impacted by a patient’s health outcome, with the 

gender gap being most pronounced among patients who experienced fatal injuries. In 

those cases, plaintiffs on behalf of female patients who died recovered significantly less 

than plaintiffs on behalf of male patients who died (p=.01). There was also a statistically 

significant gender gap in recovery amounts among plaintiffs who experienced permanent 

injury, but, as shown in Figure 3 below, that gap appears to be small. 

 

Figure 3. Average Recovery Amount by Patient Health Outcome 

 

 
 

Given the significant differences in the gender gap based on patient age and health 

outcome, four additional patient models were created to separately assess each patient 

health outcome: minor, moderate, major, and fatal injury. Minors were excluded from 

these models to avoid the masking effect seen in Model 2. The first of these models, 

labeled as Model 3 in Table 2 above, tested the effects of patient characteristics—patient 

gender, patient age, whether the practitioner was older or younger than the patient, 

whether the case settled, and type of misdiagnosis alleged—on recovery amounts for 

adult patients who experienced minor injury. In this model, patient gender was a 

significant predictor of the logpay variable, but its effect was the opposite of what was 

hypothesized: Female patients recovered more than their male counterparts who 

experienced minor injuries. Patients whose cases took longer also recovered significantly 

more than their counterparts. This model explains 19.39% of the variance in the logpay  

 

0

100000

200000

300000

400000

500000

600000

Minor Injury Moderate Injury Permanent Injury Death

U
S

D

Average Recovery Amount by Health Outcome

Female Male

p=.008

p=.004



 

39.2 COLUMBIA JOURNAL OF GENDER AND LAW                         127 

 

 

 

variable, or 19.39% of the variability in the recovery amounts received by adult plaintiffs 

with minor injuries. 

 

The second of these models, labeled as Model 4 in Table 2 above, tested the same 

variables as Model 3, but for adult patients who experienced moderate injury. In this 

model, patient gender was not a significant predictor of recovery amount. The only 

significant predictor was whether the case settled or received a final judgment, with 

patients who did not settle recovering more than their counterparts who did settle. This 

model explains 23.59% of the variance in the logpay variable, or 23.59% of the 

variability in the recovery amounts received by adult plaintiffs with moderate injuries.  

 

The third model, labeled as Model 5 in Table 2 above, tested the same variables as in 

Models 3 and 4, but for adult patients who experienced major injury. Patient gender was 

a statistically significant predictor of recovery amounts, with male patients recovering 

more than their female counterparts. Younger patients, patients who did not settle, 

patients whose cases were resolved more quickly, and patients in states with lower 

proportions of black residents also recovered significantly more than their counterparts. 

This model explains 14.41% of the variance in the logpay variable, or 14.41% of the 

variability in the recovery amounts received by adult plaintiffs with major injuries.  

 

The last of this group of models, labeled as Model 6 in Table 2 above, tested the 

same variables as Models 3, 4, and 5, but for adult patients who experienced fatal injury. 

Patient gender was a significant predictor of recovery amounts, with plaintiffs on behalf 

of male patients recovering more than plaintiffs on behalf of female patients. Younger 

patients, plaintiffs who did not settle, and plaintiffs whose cases were resolved more 

quickly also recovered significantly more than their counterparts. This model explains 

21.06% of the variance in the logpay variable, or 21.06% of the variability in the 

recovery amounts received by adult plaintiffs with fatal injuries. 

 

The upshot of Models 3 through 6 is that patient gender can have a very different 

impact among different plaintiff groups. For instance, it seems that among adult patients 

with moderate injuries, patient gender is a much less influential factor than it is for other 

patient groups. It is clear from these models that patient gender meaningfully interacts 

with age and health outcome to produce varying results for different categories of 

patients. These statistical models alone, however, do not shed light on why the gender 

gap is present for most patient groups, but absent for adults with moderate injury. These 

models also do not explain why the gender gap is reversed among some patient groups, 

like adult patients with minor injuries and minor patients. What these models do show is  

 



128 COLUMBIA JOURNAL OF GENDER AND LAW 39.2  

that patient gender does not have a consistent effect and that more research is needed to 

understand the effects of patient gender on plaintiff recovery. 

 

To better understand the gender gap, or lack thereof, among different patient groups, 

the linear regressions used to produce Models 3 through 6 were used to calculate a 

predicted value for the recovery amount variable, adjusted for all variables included in 

those models, to show the varying impact of patient gender among different categories of 

patients. Figures 4–7 below show the average recovery amounts for male and female 

adult patients in each health outcome category based on those predicted values. Based on 

this set of models, it seems that gender most significantly affects plaintiff recovery 

amounts for adults experiencing minor injury and adults who experienced major or fatal 

injury, but gender has opposite impacts for these patient groups. 

 

Figure 4. Patient Characteristics Model: Adults with Minor Injury 

 

 
 

 

 

 

 

 

 

8

8.5

9

9.5

10

10.5

11

11.5

2 0 0 4 2 0 0 6 2 0 0 8 2 0 1 0 2 0 1 2 2 0 1 4 2 0 1 6 2 0 1 8

P
re

d
ic

te
d

 V
al

u
e

Year

Pat ient  Character i s t ics  Model :  Adul ts  w i th 

Minor  Injury

Male Female



 

39.2 COLUMBIA JOURNAL OF GENDER AND LAW                         129 

 

 

 

Figure 5. Patient Characteristics Model: Adults with Moderate Injury 

 

 
 

Figure 6. Patient Characteristics Model: Adults with Major Injury 

 

 

10.8

11

11.2

11.4

11.6

11.8

12

2 0 0 4 2 0 0 6 2 0 0 8 2 0 1 0 2 0 1 2 2 0 1 4 2 0 1 6 2 0 1 8

P
re

d
ic

te
d

 V
al

u
e

Year

Pat ient  Character i s t ics  Model :  Adul ts  w i th 

Moderate  Injury

Male Female

11.45

11.65

11.85

12.05

12.25

12.45

12.65

12.85

2 0 0 4 2 0 0 6 2 0 0 8 2 0 1 0 2 0 1 2 2 0 1 4 2 0 1 6 2 0 1 8

P
re

d
ic

te
d

 V
al

u
e

Year

Pat ient  Character i s t ics  Model :  Adul ts  w i th 

Major  Injury

Male Female



130 COLUMBIA JOURNAL OF GENDER AND LAW 39.2  

Figure 7. Patient Characteristics Model: Adults with Fatal Injury 

 

 
 

3. State Characteristics Model 

 

The gender difference in average recovery amount also varies greatly by state, which 

is unsurprising given the differences in medical malpractice law from state to state. States 

vary in their contextual factors, such as population size, per capita income, the population 

density of their urban areas, and the specific features of the healthcare system—from 

legislative decisions about damages caps in medical malpractice suits to the accessibility 

of Medicaid and other health insurance. States experience varying levels of medical 

malpractice litigation, as well: New York district and state courts; Texas, Pennsylvania, 

and Maryland district courts; and New Jersey, California, Illinois, and Florida state courts 

experience the highest levels of malpractice litigation in the country.162 Interestingly, 

according to analyses of the compiled dataset in this study, three of those states—New 

Jersey, Maryland, and Florida—have some of the lowest average gender disparities in 

recovery amounts, which suggests that states with more exposure to malpractice litigation 

may be better equipped to deal with the biases which may have contributed to a 

plaintiff’s injury. Table 3 below shows the five states with the greatest gender disparities 

in average recovery amounts and the five states with the smallest gender disparities. 

 
162 See Medical Malpractice Litigation Analytics, supra note 150. 

12.2

12.3

12.4

12.5

12.6

12.7

12.8

12.9

13

2 0 0 4 2 0 0 6 2 0 0 8 2 0 1 0 2 0 1 2 2 0 1 4 2 0 1 6 2 0 1 8

P
re

d
ic

te
d

 V
al

u
e

Year

Pat ient  Character i s t ics  Model :  Adul ts  w i th 

Fatal  Injury

Male Female



 

39.2 COLUMBIA JOURNAL OF GENDER AND LAW                         131 

 

 

 

Table 3. Ten States with the Greatest and Smallest Gender Disparities in Mean 

Recovery Amounts Disadvantaging Female Patients 

 

Greatest Gender Disparities Smallest Gender Disparities 

State Gender difference in 

average recovery 

amount 

State Gender difference in 

average recovery amount 

New Hampshire $195,776.20 Louisiana $167.40 

Wyoming $144,322.60 New Jersey $1,405.90 

Connecticut $143,721.20 Maryland $2,153.90 

Colorado $139,730.10 Kansas $3,195.10 

South Dakota $106,475.90 Florida $3,206.60 

 

Still, it is unclear which contextual characteristics in these states—although 

frequency of litigation may be among them—make them more or less prone to gender 

gaps in recovery amounts. Some of the states with the smallest gender disparities, such as 

Louisiana, have very low state congeniality indices,163 suggesting that perhaps legislative 

measures restricting medical malpractice litigation have similar or equally 

disadvantageous impacts on both male and female patients/plaintiffs. However, other 

states in this category, like New Jersey, have much higher state congeniality indices, 

which indicates there must be other contextual factors at play. The following state 

characteristics models were designed to identify which state-based contextual factors are 

most influential on the gender gap in recovery amounts. 

 

The state characteristics model, labeled as Model 7 in Table 2 above, tested the 

effects of various state characteristics on recovery amounts. State characteristics include 

congeniality index, population, Gini coefficient, per capita income, historical political 

party control, percent population over age sixty-five, percent female population over age 

sixty-five, percent population lacking health insurance, death rate for female residents 

due to cardiac incidents, rate of preventable death among female residents, civil caseload 

per 100,000 residents, medical malpractice caseload per 100,000 residents, and medical 

malpractice jury trial rate. In this model, patient gender was not a significant predictor of 

recovery amount, but patients in states more amenable to malpractice suits, Republican 

 
163 See supra section III.A., discussing the congeniality index variable constructed from state statutes to 

represent how receptive a particular state’s court system is to medical malpractice lawsuits. States were given 

a score on the index based on the statute of limitations for medical malpractice suits; whether the state 

currently has, or had at any point between 2004 and 2018, a cap on damages and whether the cap is for 

noneconomic damages only or for total damages; and whether the state currently uses a locality rule, or used 

a locality rule at any point between 2004 and 2018. 



132 COLUMBIA JOURNAL OF GENDER AND LAW 39.2  

states, and states with a greater percentage of black residents all recovered significantly 

more than their counterparts. This is particularly interesting because the states which are 

least amenable to malpractice suits according to the congeniality index are states in which 

the Republican party is dominant and the states which are most amenable to malpractice 

suits are states in which the Democratic party is dominant. The relationship between a 

state’s congeniality index and dominant political party is statistically significant (p=.000), 

with Republican states scoring lower on the congeniality index. Model 7 therefore 

suggests that some of the most influential state contextual factors may have contradictory 

impacts on patient recovery amounts, as both patients in states more amenable to 

malpractice suits and patients in Republican states, which are statistically likely to be less 

amenable to malpractice suits, recovered significantly more than their counterparts. This 

sheds some light on why the gender gap disappears in this model: Because influential 

state contextual variables may have opposing effects on one another, they may also have 

mitigating or amplifying effects on the impact of patient gender. When analyzed together, 

the gender gap disappears, but it is likely that the gender gap is still significant for subsets 

of patients within states with significant contextual variables, like more amenable 

malpractice laws. For example, it is possible that there is no significant gender gap 

among all patients in a state with a high congeniality index, but there is a significant 

gender gap among black patients in the same state. This would constitute a similar 

masking effect as seen in the patient characteristics models, in which the significance of 

the gender gap depended on patient age and health outcome. However, more research is 

required to parse out the contradictory impacts of these state variables. 

 

Model 7 explains 14.19% of the variance in the logpay variable, or 14.19% of the 

differences in plaintiffs’ recovery amounts. Model 7 was used to calculate a predicted 

value for the recovery amount variable, which is adjusted for all variables included in that 

model. Figure 8 below shows the average recovery amounts for male and female adult 

patients based on those predicted values. 

 

 

 

 

 

 

 

 

 

 

 



 

39.2 COLUMBIA JOURNAL OF GENDER AND LAW                         133 

 

 

 

 Figure 8. State Characteristics Model 

 

 
 

4. Cumulative Models 

 

The last statistical model is cumulative, accounting for all the variables in the 

practitioner, patient, and state characteristics models. This model serves a comparative 

function but is also designed to assess which contextual factors have the strongest impact 

on recovery amounts. It is labeled as Model 8 in Table 2 above. According to Model 8, 

patients with worse health outcomes, patients in states with higher congeniality index 

scores, patients in Republican states, patients in states with lower proportions of the 

population aged sixty-five or older, patients in more populated states, patients in states 

with larger urban populations, patients in states with a higher proportion of black 

residents, and patients whose cases took longer recovered significantly more than their 

counterparts. Patient gender was not a significant predictor of recovery amount in this 

model, but that is unsurprising given the masking effects revealed in the patient 

characteristics models and the masking effects likely present in the state contextual 

variables model. Moreover, the patient gender variable is significantly correlated with 

other variables shown to be significant in the cumulative model, including a state’s 

proportion of black residents and a state’s political affiliation. This further indicates that, 

as patient gender has interlocking effects with these variables, the effect of patient gender 

could be masked by the interlocking effects of these variables in the cumulative model. 

10.5

10.7

10.9

11.1

11.3

11.5

11.7

11.9

12.1

12.3

12.5

2 0 1 3 2 0 1 4 2 0 1 5 2 0 1 6 2 0 1 7 2 0 1 8

P
re

d
ic

te
d

 V
al

u
e

Year

State  Character i s t ics  Model

Male Female



134 COLUMBIA JOURNAL OF GENDER AND LAW 39.2  

Table 4 below presents the statistically significant partial correlation statistics for patient 

gender and each of four other variables.  

 

Table 4. Statistically Significant Partial Correlations Between Patient Gender and 

Other Variables 

 
Variable Partial 

Corr. 

Semipartial 

Corr. 

Partial 

Corr.^2 

Semipartial 

Corr.^2 

P Value 

Female Preventable 

Death Rate 

-0.1033        -0.1010        0.0107         0.0102          0.0164 

Patient Health Outcome 0.1072         0.1049        0.0115         0.0110          0.0127 

Percent Black 0.0943         0.0921        0.0089         0.0085          0.0286 

Political Affiliation 0.0855         0.0835        0.0073         0.0070          0.0471 

 

Model 8 explains 31.82% of the variance in the logpay variable, or 31.82% of the 

differences in plaintiff recovery amounts. Table 5 below provides a summary of the 

significance, or lack thereof, of patient gender variable in each of the models included in 

this study. 

 

Table 5. Significance of Patient Gender Variable in Each Model 

 

 Model Patient Gender 

Practitioner Model 
1 0.084* 

(0.036) 

Patient Models 

2 0.006 

(0.028) 

3 -0.201⁑ 

(0.045) 

4 0.010 

(0.077) 

5 0.065* 

(0.030) 

6 0.099⁑ 

(0.015) 

State Model 
7 -0.003 

(0.087) 

Cumulative Model 
8 -0.039 

(0.042) 
 

Significance: * = p<0.05; ⁑ = p<0.01 



 

39.2 COLUMBIA JOURNAL OF GENDER AND LAW                         135 

 

 

 

5. Change Over Time 

 

Based on these statistical models, it seems that there is a significant gender gap in 

recovery amounts among some subgroups of patients and that patient gender interacts 

with other contextual variables like age, patient health outcome, and state characteristics. 

The existence of the gender gap seems to have remained present over time, as evident in 

Figures 1, 4, 6, and 7, which show the predicted variables for Models 1, 3, 5, and 6, in 

which gender was a significant factor. The stability of the gender gap may be due to the 

interplay of many significant contextual factors which may have opposing effects. Even 

if all these factors change over time, the net effect of these factors on the gender gap 

might not change very much. The stability of the gender gap may also be due to the lag 

time between advances in medical science and the infiltration of those advances into the 

medical school curriculum and subsequently into clinical practice.164 However, as seen in 

Figure 9 below, which shows the change in average recovery amount by gender over 

time, there are “spikes” in the gender gap—for example, in 2008—which may be due to 

time-sensitive factors. It is unclear why those spikes occurred; further research should 

examine changes in the gender gap over time. 

 

Figure 9. Gender Differences Recovery Amount Over Time 

 

 
 

164 See Zoë Slote Morris, Steven Wooding & Jonathan Grant, The Answer is 17 Years, What is the Question: 

Understanding Time Lags in Translational Research, 104 J. ROYAL SOC’Y MED. 510, 510 (2011) (concluding 

that it takes an average of seventeen years for health research to permeate into medical practice). 

275000

295000

315000

335000

355000

375000

395000

415000

435000

455000

2 0 0 4 2 0 0 6 2 0 0 8 2 0 1 0 2 0 1 2 2 0 1 4 2 0 1 6

U
S

D

Year

Average  Recovery Amount  By Gender  Over  T ime

Female Male



136 COLUMBIA JOURNAL OF GENDER AND LAW 39.2  

Taken as a whole, the results of the previous statistical analyses and the analysis of 

the gender gap over time show that there is, at least among some patient groups, a 

significant gendered difference in recovery amounts in diagnosis-related medical 

malpractice suits; that the gender gap is variable by state; that the gender gap is heavily 

impacted by patient age and health outcome; and that the gender gap has remained 

present over time, even with the accumulation of more knowledge about women’s health. 

 

IV. Practical Implications of the Statistical Analysis of Gender Bias in Medical 

Malpractice Lawsuits for Delayed, Wrongful, or Missed Diagnoses 

 

This study addressed five overarching questions about the effect of patient gender in 

medical malpractice cases and found: 

 

1. That there is a gendered difference in the average recovery amount for 

medical malpractice plaintiffs alleging a delayed, wrong, or missed 

diagnosis, at least among some patient groups; 

2. That the existence and size of the gender gap varies by the age of the patient; 

3. That the existence and size of the gender gap varies by the patient’s health 

outcome; 

4. That the size of the gender gap varies by state, depending on a variety of 

intersecting factors; 

5. And that the gender gap has remained present over time. 

 

Based on the analyses discussed in Part III, the most influential variables affecting 

recovery amounts for medical malpractice cases alleging missed, wrong, or delayed 

diagnoses are state contextual factors, patient gender, patient age, and patient outcome.  

 

The implications of these findings suggest that the standard of care in medical 

malpractice law disadvantages women, and particularly certain groups of women, like 

women over the age of sixty, relative to their male counterparts. These findings support 

the hypothesis that the interaction of the knowledge and trust gaps with the medical 

malpractice standard of care decreases female patients’ access to equal justice in the 

courtroom after they have experienced injustice in their medical care. While there are 

many simple changes which could be made in the healthcare context that may help 

decrease this gap, such as changing intake screening questions,165 these changes would 

 
165 For instance, it is not yet standard practice to ask about family history of autoimmune diseases, which 

disproportionately affect women, even though they tend to run in families. See DUSENBERY, DOING HARM, 

supra note 4, at 144. 

 



 

39.2 COLUMBIA JOURNAL OF GENDER AND LAW                         137 

 

 

 

not address the bulk of the knowledge gap and would not address the trust gap at all. 

Moreover, as Dr. Daniel Clauw, director of the Chronic Pain and Fatigue Research 

Center at the University of Michigan, says, “medical school curricula tend to change at a 

glacial pace,”166 which means women may be waiting an entire generation for the quality 

of care they receive to catch up with today’s medical knowledge. Other changes in the 

medical field, such as requiring a women’s health component to practitioners’ Continuing 

Medical Education, may be more effective, but also more difficult to implement. 

 

The legal system is currently protecting the lag time between medical research and its 

incorporation into medical practice and is disincentivizing change. Ensuring that the 

medical malpractice standard of care protects women patients, and particularly those who 

have experienced the negative stigma of the trust gap, may force change in the healthcare 

space and promote higher quality care for women. If there is no or a lesser penalty for 

distrusting and discounting women patients, then the gender gap in quality of care will 

continue. The medical malpractice standard of care should therefore not continue to 

protect the common practice of undervaluing women patients’ accounts of their own 

symptoms. 

 

A. Limitations of the Data 

 

It is important to note that this study assessed only recovery amounts, or the degree 

of success of patients/plaintiffs as measured by monetary recovery. In other words, all 

plaintiffs in the compiled data set used here were, to some degree, successful in 

recovering for their medical malpractice claims. That patient gender affects recovery 

amount suggests that it may also affect the likelihood of any recovery in a diagnosis-

related medical malpractice case. Recall the case of Edna Wilburn, in which a court ruled 

that practitioners’ repeated dismissal of Edna’s symptoms as psychological, though it 

delayed diagnosis of her blood disorder, did not amount to medical malpractice.167 Or 

consider the Pleasants case, in which a court ruled that a practitioner’s discounting of a 

teenage girl’s severe abdominal pain, resulting in her death, was not medical 

malpractice.168 Cases like Wilburn and Pleasants would not be captured by the present 

study because those plaintiffs did not recover at all. Their cases would not be recorded in 

a centralized database like the NPDB, as was used to create the compiled dataset for this 

 
166 Interview with Daniel Clauw by Maya Dusenbery, quoted in DUSENBERY, DOING HARM, supra note 4, at 

203; see also Morris, Wooding & Grant, supra note 164, at 510 (concluding that it takes an average of 

seventeen years for health research to permeate into medical practice). 

 
167 Wilburn v. Cleveland Psychiatric Inst., 743 N.E.2d 515, 519 (Ohio Ct. App. 2000). 

 
168 Pleasants v. Alliance Corp., 209 W. Va. 39, 41–42 (2000). 



138 COLUMBIA JOURNAL OF GENDER AND LAW 39.2  

study. Future research should attempt to fill this gap by assessing gender differences in 

success rates, not just recovery amounts, for diagnosis-related malpractice suits, 

controlling for the contextual factors assessed in this study—namely, patient age, patient 

health outcome, and state contextual factors. Future research should also aim to address 

issues of intersectionality, as aforementioned in Part I.C.4., specifically by collecting data 

on patient and practitioner race and gender, which is not collected by the NPDB, and by 

assessing which contextual factors might make the gender gap vary so much from state to 

state. Lastly, future research should analyze the text of court opinions and brief filings in 

medical malpractice cases to offer greater insight as to how the trust and knowledge gaps 

interact with the standard of care in medical malpractice cases. 

 

Future research is also necessary because the present study is limited in several 

important ways. Even if success rate data including cases like Wilburn and Pleasants was 

available, there is a funneling effect for plaintiffs who bring such suits on allegations of 

delayed, wrongful, or missed diagnoses. A correct diagnosis is generally needed to show 

that the original wrong diagnosis or the delay of diagnosis was negligent; this requires 

time and resources prior to the bringing of the lawsuit, which limits patients’ and their 

families’ ability to seek justice in the legal system for unjust treatment in the medical 

system. The ratio of plaintiffs who bring malpractice suits to patients who experience 

medical error is estimated to be between one in six and one in eight.169 Several 

demographic groups, like low income and elderly patients, may not be able to take 

advantage of medical malpractice law opportunities for recovery, as discussed in Part 

II.C.170 Future research should consider ways to incorporate these patients’ unpursued 

legal claims in an analysis of gender bias in medical malpractice law. 

 

This dataset is also limited in its ability to show intersectional effects. This dataset is 

confined to a gender binary, given that the NPDB only codes patient gender as male or 

female. The NPDB does not collect data on the race of the patient, nor the race or gender 

of the healthcare professional. These limitations should be assessed with more detailed 

data collection and analyses of these missing demographic and contextual factors. 

 

CONCLUSION 

 

The knowledge and trust gaps impact women patients’ medical malpractice cases for 

diagnosis-related issues, but the mechanics of how gender affects recovery is still unclear. 

The present study has found significant differences in recovery amounts for cases in 

 
169 See PUBLIC CITIZEN: CONGRESS WATCH, supra note 105, at 1. 

 
170 See Allen & Pierce, supra note 153. 



 

39.2 COLUMBIA JOURNAL OF GENDER AND LAW                         139 

 

 

 

which the patient is female versus cases in which the patient is male, even accounting for 

a host of contextual factors, among adult patients experiencing serious or fatal injuries 

and particularly among such patients over age sixty. The present study adds to the current 

body of research by showing that there is a gendered injustice in both the medical and the 

legal fields and the gender imbalance in each field reinforces the other. The current 

medical malpractice standard of care protects a disincentive to change the knowledge and 

trust gaps, and some policy change is likely needed in both the medical and the legal 

spheres to better protect women patients and plaintiffs in medical malpractice cases. 

 

Future research should focus on learning more about the impact of the knowledge 

and trust gaps on women patients’ legal odds as well as on disentangling the web of 

interconnected contextual factors to better understand how and when gender matters in 

medical malpractice suits. For instance, one state’s statutes and legal precedent might be 

more or less conducive than another’s to the interplay of the knowledge and trust gaps 

and the medical standard of care. Better understanding the role of patient gender in the 

medical malpractice context will inform and guide a more targeted policy change agenda 

to protect patients’ rights.