15   SA JournAl of PhySiotherAPy 2013 Vol 69 no 2 

Research

Article

MissiNG MeDical RecORDs: 
aN Obstacle tO aRchival 

sURvey-ReseaRch iN a RURal 
cOMMUNity iN sOUth aFRica

Correspondence Author:
Liezel Wegner
Department of Physiotherapy
University of the Western Cape
Private Bag X 17
Bellville 7530
South Africa
Email: liwegner@uwc.ac.za

AbstrACt:  Keeping good quality medical records is an essential yet often 
neglected part of a health-care practitioner’s workload. in South Africa, by law 
all health care facilities are required to retain medical records for a minimum of 
six years after the cessation of a patient’s treatment. In an archival survey that 
was attempted in a rural community in South Africa, only 39% of the records 
that were requested were located. The procedure that was followed in order to 
obtain the records to be included in the survey is briefly described in this paper, 
highlighting the challenges experienced in four district hospitals in this community. 
The phenomenon has serious implications not only for the quality of healthcare, 
incidence of iatrogenic injuries and the future of the health-care practitioner’s career, but it also impacts on the ability to conduct 
research to inform practice. An aspect that is not often considered is the impact of poor record keeping on the research and teaching 
component of the broader medical profession.

Key words:  meDicAl, RecoRD keePing, ARchivAl ReSeARch.

Wegner l (Msc)1

Rhoda a (PhD)1

 1 University of the Western Cape, 
  South Africa.

is it imperative to keep records but the 
quality of the records should also be 
considered. A direct correlation has been 
found between the quality of records and 
quality of care at health­care facilities in 
the Netherlands (Zegers et al., 2009). 
This finding is supported by the South 
African Medical Protections Society 
(MPS) booklet which suggests that poor 
record keeping is a significant cause of 
adverse events and iatrogenic injuries 
during hospital stay (MPS, 2012).

Inadequate care could result in  
patient dissatisfaction with the health­
services delivered. The patient could 
thus poten tially file a malpractice com­
plaint against a health­care practitioner. 
In such cases, medical records also serve 
as a medico­legal document (HPCSA, 
2008). According to the MPS casebook, 
inadequate record keeping can easily 
result in the end of a health­care prac­
titioner’s career if a complaint is made 
against the practitioner, and the medi­
cal records are insufficient to defend the 
practitioner’s case (MPS, 2012).   

iNtRODUctiON
By law adequate medical records should 
be kept by all health professionals. 
These records contain valuable infor­
mation regarding the patient’s condi­
tion and management at the health­care  
institution (Health Professionals Council 
of South Africa (HPCSA, 2008). In addi­
tion medical records are a key source of 
quantitative information in social sci­
ence research studies (HPCSA, 2008).

The purpose of keeping records is 
to ensure the continuity of care and to 
share relevant information with other 
members of the multi­disciplinary team 
(HPCSA, 2008; Claeys, 1996). Not only 

The HPCSA has clear guidelines 
on how patient information should be 
recorded and retained, as well as the 
importance of maintaining and retaining 
these records. All health­care records 
should include the patient’s general 
demographic details, records of the date 
and place of every consultation, relevant 
history, assessment, and the management 
of the patient (HPCSA, 2008; MPS, 
2000). According to the HPCSA, one of 
the reasons for the retention of medical 
records is to promote teaching and 
research (HPCSA, 2008).

In South Africa, all private and public 
health­care facilities should retain patient 
records for at least six years (HPCSA, 
2008). Records of minors should be kept 
until the age of 21 years and persons 
with mental impairments’ records must 
be kept for the duration of their lifetime. 
All records of patients who were injured 
on duty should be kept for a period of  
20 years after treatment ceased (HPCSA, 
2008). The value of records for research 
purposes should also be evaluated before 



16   SA JournAl of PhySiotherAPy 2013 Vol 69 no 2

any records can be destroyed since 
medical records are frequently used 
as a source of quantitative information  
for research studies (Zegers et al., 2009; 
Raff & James 2003). Records need to 
be legible, comprehensive and retained 
for researchers to be able to extract 
the necessary information from these 
records. In a South African survey on 
anaesthetic records done by Raff and 
James (2003), only 29.9% of the sample 
of 284 records was legible and complete. 
Missing data influences the validity and 
reliability of the findings and can result 
in missing data bias.  

This article aims to demonstrate how 
missing records / data influenced archival 
research in four district hospitals in a 
rural community in South Africa. The 
intended archival survey formed part of a 
doctoral research project. The proposed 
archival survey was developed to pro­
vide baseline information regarding 
the incidence, demographic profile and 
causes of lower limb amputations of 
patients living in a rural community 
in South Africa. The methods used to 
obtain the patient folders necessary for 
the archival survey will be described and 
the difficulties and possible reasons for 
these difficulties will be presented under 
results and discussion. 

PROceDURe OF aRchival Data 
cOllectiON
After ethical clearance to conduct 
the study was obtained from the 
ethics committee at the University 
of the Western Cape (UWC), four 
district hospitals that fall within a 
rural community in South Africa were 
conveniently selected. The records of 
all patients who received lower limb 
amputations between 1 January 2009 and 
31 December 2010 at these hospitals 
were identified as the study population. 

In order to identify the patients who 
suffered lower limb amputations, the 
theatre records at the selected hospitals 
were surveyed. Since a specific popu­
lation’s (amputees) records had to be 
retrieved, and no electronic database 
of patient’s diagnosis were available, 
theatre records were scrutinized to 
identify all patients who received lower 
limb amputations in 2009 and 2010. 
Information recorded included the 

ResUlts

“Missing” records
Between the four district hospitals 
according to the theatre records at 
Hospitals A, C and D, and the Physio­
therapy records at Hospital B, 64 patients 
received lower limb amputations during 
the research period. However, many of 
these folders had to be excluded from 
the research (see Figure 1). A total of 22 
(34%) hospital folders could be located 
or were retained by the hospitals. Twenty 
folders contained information on the 
management of the patient. The two 
folders that were excluded contained no 
information on the management of the 
patient. Thus only 20 out of 64 (31%) of 
all the records sought  could be included 
in the research project. 

hospital a
For the period between 1 January 2009 
and 31 December 2010, 25 persons 
received lower limb amputations at 
Hospital A according to the theatre 
records. Sixteen lower limb amputa tions 
were done in 2009 and nine in 2010. Of 
these 25 patients only 19 could be traced 
in the ward admissions records in order 
to establish the patients’ hospital folder 
numbers. In total ten out of 19 records 
from the ward admission book were 
retrievable from the records room less 
than three years after end of treatment 
(52.6%). Folders were filed in horizontal 
filing cabinets.

hospital b
At this hospital 10 patients were 
admitted as in­patients after lower limb 
amputations. Four of these amputees 
were admitted in 2009 and six in 2010.
Of the 10 patients identified as being 
treated in the wards of the hospital as 
in­patients (from the Physiotherapy 
records) only eight patients could be 
located in the ward admission records 
in order to trace their folder numbers. 
Only three out of the eight (37.5%) 
patients’ (for which folder numbers were 
available) records could be located. Out 
of the three folders that could be traced, 
two folders contained no information 
on the management of the patient, 
only demographic information. Out of 
10 folders initially sought, only one 

names of patients and wards that the 
patients were admitted, not the hospital 
folder numbers. Since the medical 
records are filed according to folder 
numbers, the folder number needed to 
be obtained in order to trace the folder in 
the records room. In order to determine 
patient folder numbers, the researcher 
had to use an alternative approach. Once 
the names of patients who had lower 
limb amputations in 2009 and 2010 were 
obtained from the theatre records, the 
patients’ names and date of amputation 
were used to determine the hospital in­
patient folder number from the wards’ 
admission records.  The only exception 
was for Hospital B. All amputations 
for patients serviced by this hospital 
were done at the regional hospital in the 
area since no orthopaedic surgeon was 
available at this specific district hospital 
to perform amputations during 2009 and 
2010. Information about patients with 
lower limb amputations at this hospital 
was obtained from the Physiotherapy 
records since the theatre records would 
not reflect any amputations. The Physio­
therapists at Hospital B assesses and 
manages all in­patients who received 
amputations. Once the folder number 
was obtained from the ward records, 
these names and folder numbers were 
then given to the respective hospitals’ 
record clerks to obtain the folders from 
the records room. After the folders were 
obtained, the researcher scrutinized 
the records in order to complete the 
information on the data extraction sheet 
of the archival survey.

 
ethical cONsiDeRatiONs
Ethical considerations for research 
in social sciences were adhered to. 
Permission to conduct the research 
study was obtained from the University 
of the Western Cape’s Senate Higher 
degrees Committee. Permission was 
obtained from the relevant provincial 
Department of Health, to conduct this 
study in government hospitals, as well 
as from the facility managers of the four 
district hospitals that were identified to 
be included in the study. A summary of 
the results of the research project will be 
disseminated to the various stakeholders 
involved.



17   SA JournAl of PhySiotherAPy 2013 Vol 69 no 2 

contained medical information which 
could be used in the study. The record 
clerk did now allow the researcher to 
view the records room at Hospital B.

hospital c
Twenty patients received lower limb 
amputations according to the theatre 
records, 11 in 2009 and nine in 2010. 
All twenty patients’ folder numbers 
could be located but only five out of 20 
(25%) of these folders could be traced 
in the records room. Four folders dated 
from 2009 and only one from 2010.  
At this hospital all the records in the 
records room were lying in a heap on the 
floor of the records room and not filed in 
a methodical manner.

hospital D
At Hospital D, nine patients received 
lower limb amputations. Six amputations 
were done in 2009 and three in 2010. 
Of these nine patients in­patient folder 
numbers could only be located for six of 
these patients from the ward admission 
records. Four out of nine (44%) folders 
could be retrieved from the records 
room and were available to peruse. 
Three folders were retrieved with the 
right folder number but the names did 
not match the names of the patients 
who received amputations. Folders were 
filed in horizontal filing cabinets, but  
the cabinets were too full, and folders 
were bulging out of the cabinets and 
very difficult to retrieve.

incorrect documentation
The first problem with incorrect docu­
mentation was the discrepancy between 
the theatre records and ward admis­
sion records. Routinely all patients that 
suffered an amputation will be admitted 
as an inpatient into a ward in the hospital. 
Of the 64 patients that had amputations 
according to the theatre records, only 
53 (83%) patients were admitted to the 
wards according to the ward admission 
records. 

Incorrect documentation also proved 
to be an obstacle to locating folders. Five 
out of the 64 (8%) folders were located 
according to their folder numbers but 
then presented with a different patient 
name than that of the patient requested.

Figure 1: Possible reasons for fol ders being excluded from the archival survey 
(n=64).

table 1: summary of folders obtained at each hospital.

hospital Number of 

patients 

identified to 

be included

Number 

of patient 

names 

with folder 

numbers

incomplete 

folders

Number 

of folders 

retrieved

Number 

of folders 

used

hospital a 25 19 0 10 10

hospital b 10 8 2 3 1

hospital c 20 20 0 5 5

hospital D 9 6 0 4 4

Different folder numbers
Ten out of the 64 (16%) of the patients 
that had amputations had “out­patient” 
folder numbers which could not be 
traced to a hospital folder number in 
order to obtain the folder containing the 
in­hospital management.

incomplete folders
Another hindrance in obtaining data from 
the records was incomplete records. Two 
of the records that were located only 
contained demographic information of 
the patient and no information regarding 
the patient management, and one folder 
contained some information on patient 
management, but did not state that the 
patient had an amputation at all, even 
though it was recorded in the theatre 
records and the ward admissions book.

Method of filing
Lastly, three of the four hospitals 
subjectively seemed to have a lack of 
storage space for medical records. At 
one of the hospitals records were lying 
in piles on the floor and were not filed at 
all, and at the other two hospitals filing 
shelves were overloaded with records 
and records were wedged in amongst 
each other so tightly that it was very 
difficult to remove or replace a folder 
once it was removed from the shelve.

DiscUssiON
The aim of this paper was to demon­
strate how inadequate record keeping 
can become an obstacle to research, 
since the archival survey could not be 
conducted due to an inability to retrieve 
the folders for a specific population 



18   SA JournAl of PhySiotherAPy 2013 Vol 69 no 2

group. This finding is resonated by the 
quote by Orbeta (n.d) ‘My research was 
heavily hindered by poor documen­
tation so I plan to educate the medical 
community about this issue to keep other 
valuable research from going to waste’.

Only 39% of the folders that were 
sought for use in the archival survey 
could be retrieved. Unfortunately no 
similar study could be found which 
reports on the difficulty locating folders. 
In this study only 3% (n=2) of the folders 
that were retrieved were incomplete. 
This finding compares well with the 
prospective survey that was conducted 
by Raff and James (2003) where 25% of 
the records contained no information of 
anaesthetic procedures and 45% of the 
total sample was incomplete or illegible. 

Little attention is being paid to the 
folder number when a new patient is 
admitted to the hospital. A patient might 
be seen initially as an out­patient, and 
have an out­patient folder number. If this 
patient is then admitted into the hospital 
a new folder might not be opened and 
the patient’s will continue using his out­
patient folder number. These numbers 
are very similar but the in­patient and 
out­patient folders are normally filed and 
stored separately. The ward admission 
records will then record the folder 
number but not realise or specify that it 
is an out­patient number. This folder will 
then be difficult to trace in the in­patient 
records room. 

The incorrect or non­documentation 
in the hospital ward admission records 
could be considered a major barrier 
to locating patient folders. Eleven of 
the patients, who suffered lower limb 
amputations according to the theatre 
notes, were not recorded as being 
admitted as in­patients in the wards 
admission books. No electronic records 
of patients and folder numbers were kept 
at any of these district hospitals, which 
makes locating a specific population of 
patients’ hospital folders a tedious task 
and greatly reliant on accurate manual 
record keeping. Electronic record keep­
ing could possibly ease the task of 
obtaining a specific sample of records 
for research purposes (Orbeta, n.d).

The challenge this filing system could 
potentially present is that when a patient 
came for a follow up visit as an outpatient 

after he was hospitalised and did not 
have his hospital folder number, a new 
folder would be opened for the patient. 
The health­care practitioner attending 
to this patient as an outpatient would 
then have no information regarding the 
history and previous management of this 
patient. This finding directly opposed 
the principle motive for keeping medical 
records (HPCSA, 2008). As a result one 
patient might have numerous folders 
which impedes on the limited storage 
space for medical records at health­care 
facilities, and this could directly lead 
to adverse events or iatrogenic injuries 
(MPS, 2012).

Another challenge was that even 
though the folder number might have 
been obtained the researcher would be 
given a folder with a matching folder 
number but a different patient name and 
condition. This faulty record­keeping 
could easily cause mismanagement if the 
practitioner did not check the patient’s 
name, resulting in adverse events or 
iatrogenic injury (Zegers et al., 2009; 
MPS, 2000).

Records clerks were always very busy 
servicing patients awaiting their folders 
in order to be seen by a health­care 
practitioner. The low number of records 
that could be located could subjectively 
possibly be a result of the clerks already 
being overworked and not having an 
efficient filing system which would ease 
the manual location of records (Clayes, 
1996).

Only 39% of the records that were 
sought for the intended archival survey 
in 2011 were available and accessible 
less than two years after treatment (2009 
and 2010). No statistically significant 
conclusions could be drawn due to 
the small sample size, thus affecting 
the quality of the research. These 
four hospitals did not comply with 
the HPCSA guidelines stating that all 
medical records should be kept for at 
least six years after treatment (HPCSA, 
2008). If the records are not adequately 
retained it cannot be reviewed in case 
of a query, and it cannot be utilised for 
research purposes. The records obtained 
in this specific sample also did not 
comply with the HPCSA’s specifications 
on the minimum information that should 
be included in the medical records. Three 

of the 22 folders that were retrieved had 
no information other than the patient’s 
demographic details in it. 

Medical research is vital for teaching 
purposes as well as determining common 
trends in disease and disability (Bonita, 
Beaglehole & Kjellstrom, 2006). Having 
access to medical records to review is 
imperative as these are valuable sources 
of data. The information obtained from 
medical records enables the health­care 
providers to plan and implement preven­
tative measures to improve the health of 
the overall population (Bonita, Beagle­
hole & Kjellstrom, 2006).  Secondly 
research is essential to identify problems 
in health­care service delivery to faci li­
tate improvement of health­care systems 
(HPCSA, 2008).

cONclUsiON
In conclusion, inadequate record keeping 
was found to be a major obstacle to doing 
archival research in a rural community 
in South Africa. An inability to retrieve 
the necessary medical records hindered 
the research to such an extent that the 
study had to be delayed by nearly a year 
and changed to a prospective survey. 
This change has implications for human 
resources at the hospitals, and has 
financial implications which can further 
hamper the research. 

Record keeping is the lawful respon­
sibility of every health­care practitioner, 
and retention of medical records the duty 
of all health­care facilities. Inadequate 
record keeping can compromise the 
health of the patient as well as the career 
of the health­care practitioner. It is also 
a major obstacle in archival survey 
research as were the case demonstrated 
in this paper. Lack of or poor medical 
research will in turn affect the quality 
of medical teaching and the quality of 
health­care services negatively. Accord­
ing to Pourasghar et al 2008, it is 
necessary to find ways to ensure that the 
documentation of information will be in 
a readable and retrievable format.



19   SA JournAl of PhySiotherAPy 2013 Vol 69 no 2 

ReFeReNces 

Bonita R, Beaglehole R, Kjellstrom T. (2006). 
WHO Basic epidemiology, 2nd Edition. Geneva 
Switzerland: WHO Press.

Claeys T. (1996). Medical filing, 2nd Edition. 
Delmar Cengage Learning.

HPCSA (2008). HPCSA guidelines for good 
practice in the health care professions: Guidelines 
on the keeping of patient records. Booklet 
14. Pretoria. Retrieved 15 May 2012 from:  
http://www.hpcsa.co.za/downloads/conduct_
ethics/rules/generic_ethical_rules/booklet_14_
keeping_of_patience_records.pdf

Medical protection society (2000). Making and 
keeping medical records. MPS Casebook 13 
(International). Retrieved 11 May 2012 from: 
http://www.medicalprotection.org/Default.aspx? 
DN=1843c3ce­9570­41f3­b9f1­11bfdc2138ba

Medical protection society (2012). MPS Case book 
20 (International). Retrieved on 11 May 2012 from: 
http://www.medicalprotection org/Default.aspx? 
DN=f0c10d67­a680­44a7­a599­ba1b3c7384e0

Orbeta P. (n.d). Poor medical documentation:  
Its Consequences and future considerations. Final 
report for Carnegie Mellon University. Retrieved 
12 May from: http://www.contrib.andrew.cmu.
edu/~porbeta/files/67390_final_report.pdf

Pourasghar F, Malekafzali H, Kazemi A, Ellenius J 
& Fors U. (2008). What they fill in today, may 
not be useful tomorrow: Lessons learned from 
studying Medical Records at the Women hospital 
in Tabriz, Iran. BMC Public Health, 8:139.

Raff M and James MFM (2003). An audit of 
anaesthetic record keeping. Southern African 
journal of anaesthesia & analgesia ­ July (p. 7­9). 

 
Zegers M, de Bruijne MC, Spreeuwenberg P, 
Wagner C, Groenewegen PP, van der Wal G. (2009). 
Quality of patient record keeping: an indicator 
of the quality of care? BMJ Qual Saf. Retrieved  
12 May 2012 from: http://qualitysafety.bmj.com/
content/early/2011/02/08/bmjqs.2009.038976.
abstract4343