Hrev_master The impact of C-reactive protein testing on treatment- seeking behavior and patients’ attitudes toward their care in Myanmar and Thailand Rachel C. Greer,1,2 Thomas Althaus,3,4 Sabine Dittrich,2,5,6 Christopher C. Butler,7 Phaik Yeong Cheah,1,2,8 Tri Wangrangsimakul,1,2 Frank M. Smithuis,2,9,10 Nicolas P.J. Day,1,2 Yoel Lubell1,2 1Mahidol Oxford Tropical Medicine Research Unit, Faculty of Tropical Medicine, Mahidol University, Bangkok, Thailand; 2Centre for Tropical Medicine and Global Health, Nuffield Department of Medicine, University of Oxford, Oxford, UK; 3The Department of Health Action, Monaco, Monaco; 4Monaco Scientific Centre, Monaco, Monaco; 5FIND, global alliance for diagnostic, Geneva, Switzerland; 6Deggendorf Institute of Technology, European- Campus Rottal Inn, Pfarrkirchen, Germany; 7Clinical Trials Unit, Nuffield Department of Primary Care Health Sciences, University of Oxford, Oxford, UK; 8The Ethox Centre, Nuffield Department of Population Health, University of Oxford, Oxford, UK; 9Myanmar Oxford Clinical Research Unit, Yangon, Myanmar; 10Medical Action Myanmar, Yangon, Myanmar Abstract C-reactive protein (CRP) point-of-care testing can reduce antibiotic prescribing in primary care patients with febrile and respi- ratory illness, yet little is known about its effects on treatment-seeking behavior. If patients go on to source antibiotics else- where, the impact of CRP testing will be limited. A randomized controlled trial assessed the impact of CRP testing on antibiotic prescriptions in Myanmar and Thai primary care patients with a febrile ill- ness. Here we report patients’ treatment- seeking behavior before and during the two- week study period. Self-reported antibiotic use is compared against urine antibacterial activity. Patients’ opinions towards CRP testing were evaluated. Antibiotic use before study enrolment was reported by 5.4% while antimicrobial activity was detected in 20.8% of samples tested. During the study period, 14.8% of the patients sought additional healthcare, and 4.3% sourced their own antibiotics. Neither were affected by CRP testing. Overall, patients’ satisfaction with their care and CRP testing was high. CRP testing did not affect patients’ treatment-seeking behavior during the study period whilst modestly reducing antibiotic prescriptions. CRP testing appears to be acceptable to patients and their caregivers. Introduction C-reactive protein (CRP) point of care (POC) testing can improve antibiotic pre- scribing by reducing initial antibiotic pre- scriptions for adults and children attending primary care with respiratory tract infec- tions (RTIs).1,2 The majority of RTIs are viral and do not benefit from antibiotics, but despite this, RTIs remain a common reason for an antibiotic prescription. High levels of antibiotic prescribing have been reported in Southeast Asia; situational analyses of pub- lic primary care facilities in Myanmar revealed that antibiotics were prescribed to 87% (range 73-96%) of patients with upper respiratory tract infections (URTIs), while in Thailand, 43% (20-52%) were prescribed antibiotics during 2014 and 2015.3 Thailand has been active in developing antimicrobial stewardship policies and plans, and this appears to be reducing antibiotic prescrip- tions for URTIs.3-5 Optimal use of antibi- otics is key to reducing the burden of antimicrobial resistance. In 2019, an esti- mated 254,000 deaths were attributable to bacterial resistance in Southeast Asia.6 CRP is an acute-phase protein that is raised in infection and inflammation. It can be measured at the POC using a finger prick blood test. Qualitative studies suggest that the majority of patients view the CRP POC test favorably.7-9 Less is known about its effect on treatment-seeking behavior after the initial consultation and whether patients comply with the recommendation to take or more likely not to take antibiotics. Researchers and healthcare workers have suggested that patients will go on to seek healthcare or antibiotics from other sources if they are unsatisfied with consultations using CRP POC tests.8 Whilst in research settings consultation at study sites has been largely unaffected by CRP interventions, there is a paucity of data on CRP testing’s effect on seeking healthcare and antibiotics from alternative sources.4,10-14 Trial partici- pants report conflicting views about CRP testing’s impact on future care-seeking; some report that they will re-attend to receive another test (medicalizing a self- limiting illness) while others will delay seeking care as antibiotics were not Healthcare in Low-resource Settings 2023; volume 11:11278 Correspondence: Rachel C. Greer, Mahidol Oxford Tropical Medicine Research Unit, Faculty of Tropical Medicine, Mahidol University, 3rd Floor, 60th Anniversary Chalermprakiat Building, 420/6 Ratchawithi Rd., Ratchathewi District, Bangkok 10400, Thailand. Tel.: +66.2.2036333 - Fax: +66.2.354 9169 E-mail: rachel@tropmedres.ac Key words: C reactive protein, antibiotics, AMR, point of care testing, healthcare-seeking behaviour. Contributions: YL, SD, funding acquisition; RCG, TA, SD, TW, FMS, NPJD, YL, study design; RCG, TA, YL, data analysis; YL, NPJD, PYC, CCB, super- vision; RCG, YL, first draft of the paper. All authors reviewed and edited the paper. Ethics approval: ethical approval was received from the Oxford Tropical Research Ethics Committee, the Mahidol University Faculty of Tropical Medicine Ethics Committee, the Myanmar Department of Medical Research, and the Chiangrai Provincial Public Health Office Research Ethics Committees. All participants gave their written informed consent, or assent and parental consent to join the trial. Availability of data and material: data access will be granted upon reasonable request from the Mahidol Oxford Tropical Medicine Research Unit’s Data Access Committee. Instructions and the data applica- tion form are available from: https://www.tropme- dres.ac/units/moru-bangkok/ bioethics-engage- ment/data-sharing. Conflict of interests: SD was employed by FIND the global alliance for diagnostics during the study peri- od. The other authors declare no conflict of interest; all authors confirm accuracy. Funding: this trial was funded by the Wellcome Trust Institutional Strategic Support Fund grant (105605/Z/14/Z) and Foundation for Innovative New Diagnostics (FIND) funding from the Australian Government. This research was funded in part, by the Wellcome Trust [220211]. The funders and spon- sor had no role in the study design; in the collection, analysis, and interpretation of data; in the writing of the report; and in the decision to submit the article for publication. For the purpose of Open Access, the author has applied a CC BY public copyright licence to any Author Accepted Manuscript version arising from this submission. Acknowledgments: we would like to thank all the trial participants, their caregivers, and the healthcare workers involved in this study. We would like to acknowledge the hard work of all the research nurs- es, clinicians, and clinical trial support groups at Mahidol Oxford Tropical Medicine Research Unit. We thank Dr. Daranee Intralawan and Dr. Supalert Nedsuwan from the Chiang Rai Regional Hospital for facilitating the study in the primary care units. Received for publication: 27 February 2023. Accepted for publication: 24 June 2023. This work is licensed under a Creative Commons Attribution 3.0 License (by 3.0). ©Copyright: the Author(s), 2023 Licensee PAGEPress, Italy Healthcare in Low-resource Settings 2023; 11:11278 doi:10.4081/hls.2023.11278 Publisher's note: all claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher. [page 42] [Healthcare in Low-resource Settings 2023; 11:11278] needed.7,12 Patient-reported antibiotic use can be difficult to assess and validate due to a lack of awareness or understanding of antibiotics and other medications being taken, as well as poor adherence to treat- ment and recall times. Measuring urine antibacterial activity is one way to verify whether antibiotics are being taken current- ly.10,15-17 We conducted a randomized con- trolled trial (RCT) to evaluate CRP-guided antibiotic prescribing for patients attending primary care with an acute febrile illness. The primary outcomes have been reported previously.4 In summary, a modest reduc- tion (39% vs. 34%) in antibiotic prescribing was seen in the intervention arm using a CRP cut-off of 40mg/L compared with the control arm (aOR 0.80, 95% CI 0.65-0.98). Patients with a high CRP level were more likely to receive an antibiotic and those with a low CRP were less likely to receive an antibiotic in the intervention arms com- pared to the control arm. Clinical outcomes were not affected.4 In this paper, we describe the secondary outcomes of patients’ treatment-seeking behavior (healthcare and antibiotics) before and during the two-week study period and compare self-reported antibiotic use against urine antibacterial activity. We explore patients’ and their caregivers’ views toward CRP POC testing. Materials and Methods We conducted a multicentre, open–label RCT in Myanmar and Thailand. The trial design details have been reported previous- ly.4 In brief, we recruited patients aged 1 year or older attending primary care with a documented fever (>37.5°C) or history of fever in the last 2 weeks. Patients were indi- vidually randomized 1:1:1 into intervention arm A (CRP cut-off of 20mg/L), interven- tion arm B (CRP cut-off of 40mg/L), or the control arm (standard care). These CRP cut- offs were based on reported CRP levels in Southeast Asian febrile patients and recent RCTs on POC CRP testing in primary care. Prior antibiotic use did not prevent partici- pation.4 Healthcare workers were advised that patients with a low CRP result (defined by the intervention arm’s threshold) were unlikely to benefit from antibiotics while those with a high CRP were more likely to benefit from antibiotics. All patients were followed up on days 5 and 14. Urine sam- ples were collected on day 0 and day 5. Opinions towards their care and CRP test- ing were ascertained by the researcher using close-ended questions on day 14. Study sites The study sites in Myanmar included three not-for-profit clinics which provide primary healthcare for marginalized people and one government outpatient department. All patients were treated for free by doctors. The Thai study sites were six government- run primary care units that provide univer- sal health care and medication to Thai citi- zens for a nominal fee. They are usually staffed by nurses and public health officers. In both Myanmar and Thailand, antibiotics are also available from multiple sources, such as pharmacies and village shops with- out a prescription. Laboratory procedures Urine antibacterial activity was tested at the Mahidol Oxford Tropical Medicine Research Unit (MORU) laboratory in Bangkok, Thailand. The reference organ- ism, Bacillus stearothermophilus (ATCC 7953) was plated on Mueller Hinton agar. Urine samples were thawed and then 3 μL samples were pipetted onto a blank filter paper noting the disc position. Plates were incubated aerobically at 56°C for 18 to 24 hours. If an inhibitory zone was seen around the urine sample then antibacterial activity was declared.16 Samples were tested in duplicate and divergent results were repeat- ed. All urine samples collected on day 5 were tested for antibacterial activity but only a subset of day 0 urine samples (409/2,292, 17.8%) were tested due to resource constraints. CRP levels were assessed using the NycoCard II Reader, Axis-Shield, Oslo, Norway. Capillary blood samples were test- ed at point-of-care for intervention patients whereas for control patients venous samples were retrospectively tested in MORU’s local laboratories.18 Statistical analysis Categorical data were summarised using counts and percentages, and com- pared using χ2 tests. Mann-Whitney U tests were used to compare scores without nor- mal distribution. Logistic regression models were used to evaluate indicators of treat- ment-seeking behavior during the study, with the study sites fitted as random effects. Univariate analyses of the potential indica- tors of treatment-seeking behavior were performed and significant variables (p<0.05) were added to multivariable anal- yses. Agreement between patient-reported antibiotic use and urine antibacterial activi- ty was assessed using the kappa statistic. Patients’ consultation experience scores were created using the sum of responses to questions 2, 3, 4, 8, and 9 (Table 1). Responses were recoded so that positive answers received 1 point, neutral answers 0 points, and negative responses -1 point.18 Results Health-seeking and antibiotic use before enrolment The RCT enrolled 2,410 patients with an acute fever or history of fever presenting to primary care in Myanmar and Thailand between 2016 and 2017.4 Over half of the patients (1,372/2,408, 57%) had sought healthcare in the two weeks before study enrolment, most frequently from pharma- cies (53.8%) and clinics (22.1%). Prior care was more likely to have been sought by patients in the Myanmar facilities, as com- pared with those in the Thai facilities (74.9% vs 38.4%, p<0.001), and when the patient was an adult as compared with chil- dren (61.5% vs 52.4%, p<0.001). New medication had been taken by 1,732/2,409 (71.9%) of the patients in the 2 weeks before study enrolment; of these, 367 (21.2%) had taken at least one unknown medication. Antibiotics had been knowing- ly taken by 130/2,409 (5.4%). Sources of antibiotics include clinics (81/126, 62.3%), pharmacies (30, 23.1%), hospitals (6, 4.6%), natural healers (5, 3.9%), household supplies (3, 2.3%), street vendors (1, 0.8%) and unknown (4, 3.1%). A minority of those who had sought healthcare reported taking antibiotics (127/1,372, 9.3%). Prior antibi- otic use did not vary between Myanmar and Thai patients or adults and children (p=0.347 and 0.223, respectively). Antibacterial activity was found in 85/409 (20.8%) of the urine samples tested at enrolment. The agreement between reported antibiotic use and urine antibacte- rial activity was 81.2% (kappa = 0.21). Of the 409 patients, 22 reported antibiotic use in the 48 hours before the test, of whom 15 were positive and 7 were negative for antibacterial activity, while 70/85 (82.4%) of the patients with urine antibacterial activ- ity did not report antibiotic use (Figure 1). In those who were taking an unknown med- ication, 29/61 (47.5%) of the urine samples were positive for antibacterial activity.18 Health-seeking and antibiotic use after enrolment Antibiotics were prescribed at enrol- ment to 515/1,593 (32.3%) of the patients in the CRP intervention arms compared to 297/799 (37.2%, p=0.018) in the control arm. This reduction in prescribing was pri- marily due to a reduction in Myanmar adults.4 During the study period, healthcare was sought by 339/2,294 (14.8%) of the Article [Healthcare in Low-resource Settings 2023; 11:11178] [page 43] patients (from any source or facility other than the study follow-up visits). There was no difference between those in the CRP intervention arms and the control arm (p=0.552, Supplementary Table 1). In the multivariable analysis, significantly less care was sought during the study by Thai patients and those who had received an antibiotic at enrolment. Significantly more care was sought by those who had sought care before the study, presented with a doc- umented fever, higher self-reported symp- tom severity, higher CRP results, and those diagnosed with an unspecified acute viral or dual infection compared to those with RTIs (Table 2). Antibiotics were prescribed to 110/2,311 (4.8%) of the patients on day 5 and 15/2,317 (0.7%) on day 14. In addition, 95/2,206 (4.3%) of the patients sourced their antibiotics, approximately a third of the 254 patients seeking care elsewhere; an additional 79 patients received an unknown medication. There was no difference between those seeking antibiotics in the CRP intervention or control arms. The only significant variable in the univariate analy- ses for seeking antibiotics during the study was having a higher CRP result at enrol- ment, p=0.002 (Supplementary Table 1). On day 5, urine antibacterial activity was found in 521/2,065 (25.2%) of the sam- ples (Figure 1). The overall agreement between patient-reported antibiotic use and urine antibacterial activity was 77.4% (kappa=0.46). In the preceding 48 hours, 641/2,065 (31.0%) patients reported antibi- otic use; 352 (54.9%) samples were positive and 289 (45.1%) were negative. In those with urine antibacterial activity, 352/521 (67.6%) patients had reported antibiotic use, while 155 (29.8%) reported no antibi- otic use. Most (77.6%) patients reported knowing whether they had been prescribed an antibiotic at enrolment; the rest were unsure when asked on day 14. Adherence to antibiotic courses was reported by 687/829 (86.7%) of the patients.18 Patients’ and caregivers’ opinions and attitudes toward the consulta- tion and CRP testing On day 14, all patients were asked about their care, and those in the intervention arms were asked additional questions about CRP testing. Half of the patients answered these questions themselves while the other half were answered by their parents or guardians. Overall satisfaction with the care received was very high (Table 1). There were no differences between the interven- tion and control arms in terms of consulta- tion scores (p=0.980), an adequate explana- tion of the treatment (p=0.966), or agree- ment with the treatment (p=0.864). Thai patients rated each of these higher than Myanmar patients (p<0.001). Patients who sought further healthcare during the study scored lower for their con- sultation experience (p<0.001), an adequate explanation of the treatment (p=0.007), and agreement with their treatment (p=0.006) than those who did not. Patients who sourced antibiotics during the study had similar consultation scores (p=0.313) and an adequate explanation of the treatment (p=0.847) but reported less agreement with treatment (p=0.001). Agreement with the treatment was also lower in those not pre- scribed an antibiotic at enrolment compared to those who were (p=0.033), however over 80% agreed with the antibiotic prescribing decision. In total, 67.3% of the patients reported receiving enough explanation to understand their treatment. While in the intervention arms, 61.1% felt the objective of the CRP test was clear and 56.6% reported that the test results were explained in a way that they understood. The majority of interven- tion patients wanted the CRP test to be used again, felt more confident whether antibi- otics were needed, and that it improved their quality of care.18 Discussion Following a modest reduction in pre- scribing after the first presentation, CRP POC testing did not affect patients’ treat- ment-seeking behavior during the two-week study period. Patients expressed positive opinions towards CRP testing and its use in future consultations. Studies from Asia and Europe have reported no difference in re- attendances between patients in CRP and control arms.4,10-14,19 Our study goes further by showing no difference in the numbers of patients seeking additional healthcare or antibiotics in the two weeks following first attendance at the study facility. It is encour- aging that despite the relatively low antibi- otic prescribing in the control and interven- tion arms, less than 5% of the patients went on to source their antibiotics. Taken togeth- er with the reported high adherence to antibiotic courses this should encourage healthcare workers and policymakers that most patients will comply with antibiotic treatment plans, even when antibiotics can be sought from other sources. Patients reported high levels of satisfaction with their care. This is consistent with other stud- ies on CRP POC testing.10-12 Our study raises concerns about unknown medication use. A fifth of those taking a new medication before enrolment did not know what they were taking. Even within the context of a trial focusing on antibiotic use a quarter did not know if they were prescribed an antibiotic as part of the study. Some of this uncertainty about antibi- otic use may be explained by the multiple terms used for antibiotics in Thailand and the lack of a formal word for antibiotics in Myanmar.20,21 This uncertainty is likely to be reflected in the differences between reported antibiotic use and urine antibacte- rial activity and is consistent with other studies that found lower levels of reported antibiotic use compared to urine antibacteri- al activity.15,16,22 Another reason for this dis- crepancy may be environmental exposure to antibiotics, foods or chemicals with antibac- terial activity.16,23 False-negative results may have been caused by non-adherence to antibiotics, extra-renal antibiotic excretion, reduced test sensitivity due to one reference organism being used, and the freezing and thawing of urine samples.15-18,22 Moving for- ward patients need to be aware of their antibiotic use if they are to be involved in strategies to optimize antibiotic use. Patients’ understanding of the CRP test could be improved and may help to Article Figure 1. Venn diagrams to show day 0 and day 5 urine antibacterial activity and reported antibiotic use. Adapted from Greer 2022.18 [page 44] [Healthcare in Low-resource Settings 2023; 11:11278] Article Table 1. Patients’ and caregivers' opinions and attitudes towards the consultation and CRP POC testing, by country and intervention. Adapted from Greer 2022.18 Patients’ and caregivers’ opinions & attitudes Agree Neutral Disagree n (%) n (%) n (%) I think that the healthcare worker’s decision to prescribe or not to prescribe an antibiotic for my treatment was correct (Q 2) Intervention arms (N = 1,377) 1,113 (80.8) 241 (17.5) 23 (1.7) Control arm (N = 691) 556 (80.5) 125 (18.1) 10 (1.5) Thailand (N = 1,172) 1,107 (94.5) 49 (4.2) 16 (1.4) Myanmar (N = 896) 562 (62.7) 317 (35.4) 17 (1.9) I did not get enough explanation to understand the treatment (Q 3) Intervention arms (N = 1,448) 79 (5.5) 394 (27.2) 975 (67.3) Control arm (N = 725) 37 (5.1) 200 (27.6) 488 (67.3) Thailand (N = 1,173) 54 (4.6) 269 (22.9) 850 (72.5) Myanmar (N = 1,000) 62 (6.2) 325 (32.5) 613 (61.3) I felt that the consultation was too fast (Q 4) Intervention arms (N = 1,451) 335 (23.1) 254 (17.5) 862 (59.4) Control arm (N = 726) 155 (21.4) 123 (16.9) 448 (61.7) Thailand (N = 1,174) 394 (33.6) 36 (3.1) 744 (63.4) Myanmar (N =1,003) 96 (9.6) 341 (34.0) 566 (56.4) I fully understood the instructions for taking the prescribed antibiotic (including when, how much, how often, and how long I have to take the medication) (Q 5) Intervention arms (N = 407) 388 (95.3) 15 (3.7) 4 (1.0) Control arm (N = 211) 195 (92.4) 14 (6.6) 2 (1.0) Thailand (N = 353) 343 (97.2) 8 (2.3) 2 (0.6) Myanmar (N = 265) 240 (90.6) 21 (7.9) 4 (1.5) It is too much effort to come to the health center for the treatment that I received (Q 8) Intervention arms (N = 1,461) 107 (7.3) 212 (14.5) 1,142 (78.2) Control arm (N = 732) 58 (7.9) 101 (13.8) 573 (78.3) Thailand (N = 1,173) 32 (2.7) 30 (2.6) 1,111 (94.7) Myanmar (N = 1,020) 133 (13.0) 283 (27.8) 604 (59.2) Overall, I am satisfied with my care (Q 9) Intervention arms (N = 1,464) 1,429 (97.6) 33 (2.3) 2 (0.1) Control arm (N = 730) 709 (97.1) 19 (2.6) 2 (0.3) Thailand (N = 1,173) 1,155 (98.5) 16 (1.4) 2 (0.2) Myanmar (N = 1,021) 983 (96.3) 36 (3.5) 2 (0.2) Intervention arms only The objective of the finger-prick CRP test is not clear to me (Q 6) All (N = 1,453) 64 (4.4) 502 (34.6) 887 (61.1) Thailand (N = 776) 31 (4.0) 292 (37.6) 453 (58.4) Myanmar (N = 677) 33 (4.9) 210 (31.0) 434 (64.1) The finger-prick test for CRP is painless (Q 7) All (N = 1,450) 998 (68.8) 222 (15.3) 230 (15.9) Thailand (N = 777) 672 (86.5) 36 (4.6) 69 (8.9) Myanmar (N = 673) 326 (48.4) 186 (27.6) 161 (23.9) Yes Do not know No Did the health worker explain the finger-prick test results to you in a way that you understood? (Q 10) All (N = 1,450) 821 (56.6) 299 (20.6) 330 (22.8) Thailand (N = 774) 435 (56.2) 194 (25.1) 145 (18.7) Myanmar (N = 676) 386 (57.1) 105 (15.5) 185 (27.4) Would you like the health worker to use the finger-prick test for CRP again the next time you have an illness? (Q 14) All (N = 1,461) 1,329 (91.0) 103 (7.1) 29 (2.0) Thailand (N = 778) 763 (98.1) 12 (1.5) 3 (0.4) Myanmar (N = 683) 566 (82.9) 91 (13.3) 26 (3.8) Did the health worker seem to base his/her treatment decision on the test results? (Q 12) All (N = 1,443) 782 (54.2) 557 (38.6) 104 (7.2) Thailand (N = 774) 492 (63.6) 273 (35.3) 9 (1.2) Myanmar (N = 669) 290 (43.4) 284 (42.5) 95 (14.2) Continued on the next page. [Healthcare in Low-resource Settings 2023; 11:11278] [page 45] increase the impact of CRP testing and patients’ agreement with their antibiotic treatment. The patients who did seek addi- tional healthcare during the study had lower consultation experience scores, less adequate explanation, and less agreement with their treatment. Further work is required to explore how these areas could be addressed in future interventions. Special focus needs to be given to patients who are not prescribed an antibiotic, espe- cially when antibiotics are expected and this is the prescribing norm. This manuscript adds detailed treat- ment-seeking behavior to the results of our CRP POC RCT. Combined with the urine antibacterial activity data and patient’s opinions towards CRP POC testing this pro- vides a more holistic review of the patient’s acceptance of CRP POC testing, in the con- text of two low-and-middle-income coun- tries. There are, however, several limita- tions to our study; the effect of CRP testing on treatment-seeking behavior may have differed if the intervention had had a larger impact on antibiotic prescribing. Patients’ satisfaction with their care may have been influenced by the study design, including the follow-up visits. Opinions towards CRP testing were assessed using close-ended questions which cannot give as detailed or nuanced answers as qualitative methods. However, as part of our wider work patients’ views were explored using semi- structured interviews and their opinions were widely positive.8 Due to resource con- straints we were unable to test all the enrol- ment urine samples for antibacterial activity and were only able to use one reference organism; this may have led to an underes- timation of urine antibacterial activity. Our study sites were limited to government and not-for-profit-run primary care clinics so the results may not be generalizable to other facilities. Article Table 1. Continued from previous page. Patients’ and caregivers’ opinions & attitudes Too much Enough/ adequately Not enough If so: Do you think the health worker relied too much, enough, or not enough on the test results when he/she made the treatment decision? (Q 12a) All (N = 778) 192 (24.7) 580 (74.6) 6 (0.8) Thailand (N = 491) 181 (36.9) 309 (62.9) 1 (0.2) Myanmar (N = 287) 11 (3.8) 271 (94.4) 5 (1.7) More Neither more Less confident nor less confident confident Did the finger-prick test make you feel more or less confident that antibiotics are needed / not needed for your illness? (Q 11) All (N = 1,432) 1,201 (83.9) 225 (15.7) 6 (0.4) Thailand (N = 776) 738 (95.1) 37 (4.8) 1 (0.1) Myanmar (N =656) 463 (70.6) 188 (28.7) 5 (0.8) Improves No difference, unsure Worsens Do you feel that the finger-prick test for CRP improves or worsens the quality of the care you receive? (Q 13) All (N = 1,446) 1,281 (88.6) 165 (11.4) 0 Thailand (N = 778) 753 (96.8) 25 (3.2) 0 Myanmar (N = 668) 528 (79.0) 140 (21.0) 0 Table 2. Multivariable logistic regression of variables associated with seeking healthcare during the RCT. Adapted from Greer 2022.18 Variable Additional healthcare sought during the study period aOR* (95% CI) P value Country Myanmar patients Reference Thai patients 0.43 (0.23 to 0.81) 0.008 Sought healthcare before enrolment 1.47 (1.07 to 2.01) 0.016 Documented fever at enrolment 1.75 (1.31 to 2.35) <0.001 Self-reported symptom severity score (1 point increase) 1.81 (1.33 to 2.46) <0.001 Diagnosis at enrolment# RTIs Reference Other infections 1.22 (0.70 to 2.12) 0.480 Acute viral infections (unspecified) 1.71 (1.12 to 2.63) 0.014 Dual infection 1.82 (1.04 to 3.18) 0.037 CRP level at enrolment (1 mg/L increase) 1.01 (1.00 to 1.01) 0.001 Antibiotics prescribed at enrolment 0.52 (0.37 to 0.73) <0.001 *The study site was added as a random effect. #Other infections include all non-RTIs affecting other systems such as gastrointestinal and skin infections. Acute viral infection was a common diagnosis made in Myanmar alongside RTIs, common symptoms included cough and runny nose but some patients had fever as the sole symptom. Dual infections include a diagnosis from two of the diagnosis categories. [page 46] [Healthcare in Low-resource Settings 2023; 11:11278] [Healthcare in Low-resource Settings 2023; 11:11278] [page 47] Conclusions The use of CRP POC testing has been shown to improve healthcare workers’ antibiotic prescribing practices. Here it was shown that CRP testing was widely accept- able to primary care patients in Myanmar and Thailand, without affecting subsequent treatment-seeking behavior. Encouragingly, the vast majority of antibiotics obtained outside the study facilities came from for- mal sources, such as pharmacies and clin- ics, even though in Myanmar and Thailand antibiotics are widely available from infor- mal providers, such as natural healers or street vendors. These formal providers could be easier sites to target antimicrobial stewardship interventions. Healthcare workers should communicate to patients when they are prescribing antibiotics and if they are not the reasons why antibiotics are not required. References 1. Verbakel JY, Lee JJ, Goyder C, et al. Impact of point-of-care C reactive pro- tein in ambulatory care: a systematic review and meta-analysis. BMJ Open 2019;9:e025036-e36. 2. Smedemark SA, Aabenhus R, Llor C, et al. Biomarkers as point-of-care tests to guide the prescription of antibiotics in people with acute respiratory infections in primary care. Cochrane Database Syst Rev 2022;10:Cd010130. 3. Holloway KA, Kotwani A, Batmanabane G, et al. Antibiotic use in South East Asia and policies to promote appropriate use: reports from country situational analyses. BMJ 2017;358:j2291. 4. Althaus T, Greer RC, Swe MMM, et al. Effect of point-of-care C-reactive pro- tein testing on antibiotic prescription in febrile patients attending primary care in Thailand and Myanmar: an open- label, randomised, controlled trial. Lancet Glob Health 2019;7:e119-e31. 5. Thai Ministry of Public Health. HDC service report on the service plan for rational drug use 2020. Accessed 17/07/2020 2020. Available from: https://hdcservice.moph.go.th/hdc/repo rts/report.php?source=pformated/for- mat1.php&cat_id=03b912ab9ccb4c072 80a89bf05e5900e&id=d1ccec314e928 75acb5142769eb479a2 6. Murray CJL, Ikuta KS, Sharara F, et al. Global burden of bacterial antimicro- bial resistance in 2019: a systematic analysis. Lancet 2022;399:629-55. 7. Tonkin-Crine S, Anthierens S, Francis NA, et al. Exploring patients’ views of primary care consultations with con- trasting interventions for acute cough: a six-country European qualitative study. NPJ Prim Care Respir Med 2014;24:1 4026. 8. Haenssgen MJ, Charoenboon N, Althaus T, et al. The social role of C- reactive protein point-of-care testing to guide antibiotic prescription in Northern Thailand. Soc Sci Med 2018;202:1-12. 9. Van den Bruel A, Jones C, Thompson M, et al. C-reactive protein point-of- care testing in acutely ill children: a mixed methods study in primary care. Arch Dis Child 2016;101:382. 10. Do NT, Ta NT, Tran NT, et al. Point-of- care C-reactive protein testing to reduce inappropriate use of antibiotics for non- severe acute respiratory infections in Vietnamese primary health care: a ran- domised controlled trial. Lancet Glob Health 2016;4:e633-41. 11. Cals J, Butler C, Hopstaken R, et al. Effect of point of care testing for C reactive protein and training in commu- nication skills on antibiotic use in lower respiratory tract infections: cluster ran- domised trial. BMJ 2009;338:1374. 12. Cals JWL, Schot MJC, de Jong SAM, et al. Point-of-Care C-Reactive Protein Testing and Antibiotic Prescribing for Respiratory Tract Infections: A Randomized Controlled Trial. Ann Fam Med 2010;8:124-33. 13. Little P, Stuart B, Francis N, et al. Effects of internet-based training on antibiotic prescribing rates for acute respiratory-tract infections: a multina- tional, cluster, randomised, factorial, controlled trial. Lancet 2013;382:1175- 82. 14. Schot MJ, Van den Bruel A, Broekhuizen BD, et al. Point-of-care C- reactive protein to assist in primary care management of children with suspected non-serious lower respiratory tract infection: a randomised controlled trial. BJGP Open 2018;2:bjgpopen 18X101 600-bjgpopen18X00. 15. Emary KR, Carter MJ, Pol S, et al. Urinary antibiotic activity in paediatric patients attending an outpatient depart- ment in north-western Cambodia. Trop Med Int Health 2015;20:24-8. 16. Khennavong M, Davone V, Vongsouvath M, et al. Urine antibiotic activity in patients presenting to hospi- tals in Laos: implications for worsening antibiotic resistance. Am J Trop Med Hyg 2011;85:295-302. 17. Liu Y-C, Huang W-K, Huang T-S, et al. Detection of antimicrobial activity in urine for epidemiologic studies of antibiotic use. J Clin Epidemiol 1999; 52:539-45. 18. Greer RC. Evaluation of C-Reactive Protein Point of Care Testing, and Associated Research Challenges, to Improve the Quality of Antibiotic Prescribing in the Community in Northern Thailand. PhD thesis The Open University, 2022. 19. Andreeva E, Melbye H. Usefulness of C-reactive protein testing in acute cough/respiratory tract infection: an open cluster-randomized clinical trial with C-reactive protein testing in the intervention group. BMC Fam Pract 2014;15:80. 20. Haenssgen MJ, Charoenboon N, Zanello G, et al. Antibiotic knowledge, attitudes and practices: new insights from cross-sectional rural health behav- iour surveys in low-income and middle- income South-East Asia. BMJ Open 2019;9:e028224. 21. Miyano S, Htoon TT, Nozaki I, et al. Public knowledge, practices, and awareness of antibiotics and antibiotic resistance in Myanmar: The first nation- al mobile phone panel survey. PLoS One 2022;17:e0273380. 22. Liu Y-C, Huang W-K, Huang T-S, et al. Extent of antibiotic use in Taiwan shown by antimicrobial activity in urine. Lancet 1999;354:1360. 23. Wang H, Wang B, Zhao Q, et al. Antibiotic body burden of Chinese school children: a multisite biomonitor- ing-based study. Environ Sci Technol 2015;49:5070-9. Article Online Supplementary Materials Table S1. Univariate analyses of variables for seeking healthcare and antibiotics during the RCT study period.