Cancer of the cervix is the fourth most common cancer among women worldwide, with an estimated 569,847 new cases and 311,365 deaths occurring in 2018 1. Women living in low- and middle-in-come countries (LMICs) account for a disproportionate number of cervical cancer cases and deaths, and cervical cancer is the most commonly diagnosed cancer and the leading cause of cancer death among women in Bolivia 1. Bolivia's Ministry of Health (MoH) strategies to reduce cervical cancer incidence and mortality include human papillomavirus (HPV) vaccination of girls aged 1012 years (introduced in 2016) and a nationwide scale-up of cervical cancer screening 2. MoH protocols outline cervical cancer screening using Papanicolaou (Pap) test exams or visual inspections with acetic acid (VIA) 3. Al-though VIA is seen as a promising screening method for Bolivia, it has not been rolled out nationally 4,5. National guides also mention HPV self-sampling as an effective screening method, but it is still new to Bolivia 6.
Despite these advances, cervical cancer screening coverage in Bolivia remains low. The most recent comprehensive Bolivia Demographic Health Survey conducted in 2008 reports 33.3% cervical cancer screening coverage for women aged 15-49 5,6, far below the new WHO target of 70% coverage 7. The poor quality of cervical cancer services may be an important factor in suppressing screening and treatment rates. Bolivian patients have reported shortages of providers, unfriendly providers, nonfunctioning equipment, facilities with poor hygiene, poor continuity of care, long delays for services, and lost Pap test results, especially at public health facilities 8. Approximately 50% to 80% of women who undergo cervical cancer screening in Bolivia are lost to follow-up 8. Inadequate information and tracking systems, a lack of coordination between providers for diagnosis and treatment, and weak logistics systems for transporting tests to labs and communicating results to providers and women are significant barriers to Pap screening programs in Bolivia 8. Many providers are unaware of guidelines and receive little training on new cervical cancer diagnostic techniques 8. As a result of low coverage and poor quality screenings and follow-up, fewer than 20% of eligible women in Bolivia receive treatment for precancerous cervical lesions 8,9.
Bolivia needs to assess its current cervical cancer programs to develop new policies for addressing this situation, but evaluating the quality of cervical cancer screenings in LMICs is challenging. Methods commonly used in cervical cancer health services research include analyzing medical records and other secondary data, patient and provider surveys, and direct provider observations 10. However, these methods have shortcomings. Collecting data from medical records in LMICs is problematic as these records are often patchy and charting minimal 11. Surveys can produce biased results, and providers often alter their behavior when under observation11-13. Assessing a health facility's structural measures, such as medical equipment, provider qualifications, and checklists, is not a reliable proxy for quality of care 14. Improved cervical cancer screening evaluation methods are needed in LMICs to inform policymakers.
The mystery patient method may be an important tool for addressing these limitations and improving Bolivia's cervical cancer screening quality assessments. Mystery patients, also known as simulated patients, are trained community members who visit health facilities in the assumed roles of patients and then complete a debriefing questionnaire on their experiences 15. Mystery patients are considered the gold standard for assessing clinical practice in high-income countries (HICs) 11,14,16,17, and have recently been used in LMIC settings 14 to assess barriers to care for sensitive services such as family planning, abortion, and mental health assessments 18-21. They have been particularly successful at documenting communication problems such as condescending attitudes, stigma, and religious counseling that patients face in seeking sensitive services 22. Mystery patients in LMICs have been primarily used to gather data on non-invasive practices such as counseling or dispensing medications. Little is known about their efficacy in assessing the quality of more invasive procedures such as cervical cancer screening 23.
Logistical and ethical concerns have limited the use of mystery patients to evaluate health services in LMICs. Many Institutional Review Boards (IRBs) will not approve research where mystery patients collect data in potentially dangerous situations where they might acquire an infection from a procedure such as an injection with a dirty needle. Guidelines for mystery patients recommend not undergoing medical exams or invasive procedures (such as gynecological exams) during visits 14,15,24. In a recent study that assessed the quality of health care for patients in Kenya, mystery patients were coached to avoid taking medications or undergoing blood tests, injections, X-rays and other invasive procedures during their encounters 11.
Another concern with this method is that the mystery patients could waste providers' time or be detected by providers during the data collection 11. In both HICs and LMICs, there have been questions about the need for obtaining informed consent from the health providers un-der study by mystery patients. Ethical analyses of mystery patient methods have concluded that mystery patients collecting data without the consent of the health professionals working in that system can be justified when the risks are minimal, and the research has social value 25,26. These reports suggest that IRBs could grant waivers related to informing providers about mystery patients.
Despite ethical concerns, it has been determined that measuring the quality of clinical practice using mystery patients in LMICs can be safe and effective 11. In order to ethically use mystery patients, potential ethical issues are determined and addressed early in the research design process and highlighted in IRB documents 26.
Given the potential of the mystery patient evaluation method and the need for rigorous quality assessments during the planned scale-up of cervical screening services in Bolivia, we conducted a pilot study to determine whether we could develop a mystery patient protocol that addressed logistical and ethical concerns while collecting useful quality of care information. Our objective was to assess the feasibility of using the mystery patient method to evaluate the quality of cervical cancer screening in four health facilities in rural Bolivia.
METHODS
Sample and recruitment
In coordination with local health authorities, the researchers purposively selected a rural Bolivian province, with a population of approximately 27,000 people, where health officials reported low Pap test coverage and wanted to investigate if the quality of services could explain low uptake. Together with health and social authorities, the researchers chose four communities within the province. Included in the sample was an indigenous community living far from the provincial capital.
Data collection instrument
The mystery patient debriefing questionnaire was developed using Bolivian national cervical cancer standards and international reproductive health quality improvement frameworks so evaluation results could be measured against known benchmarks 24. Drawing on clinical content and quality of care attributes, essential elements in the debriefing questionnaire included: the health facility environment, privacy and confidentiality concerns, a holistic rights-based approach with continuity of care, interpersonal communication and education, health worker competency with safe/good practice, respect for dignity and comfort, and follow-up services 2,15,27-29. Facility-related questions covered facility location, cleanliness, and access to information. Service questions captured issues related to staff/provider attitudes, fees, respect, good practice, waiting times, communication, information, and follow-up. We also integrated cross-cutting themes on women's reproductive and sexual rights related to cervical cancer services into the questionnaire.
The initial mystery patient debriefing questionnaire, which had 56 questions, was tested with ten local com-munity women using cognitive interviewing verbal probing techniques 30. Cognitive interviewing revealed that some women, especially older women, did not comprehend many terms and misunderstood technical phrases. Revisions included deleting unnecessary questions, expanding the questionnaire to 68 items to capture more laboratory results and follow-up information, and replacing technical terms with familiar phrases. For example, the term "confidential" was changed to "secret."
Mystery patient recruitment and orientation
As part of a larger study, we held focus groups about cervical cancer knowledge, attitudes, and practices with re-productive-age women. After the focus groups, women who already had their Pap test exams scheduled were invited to participate in this pilot study as mystery patients. Two women from each of the four communities (n=8) were selected to be mystery patients. The participant requirements and conditions were explained to the women during the informed consent process. A nurse on the research team, fluent in the local indigenous language, oriented the women to the debriefing questionnaire. This nurse was not working at the facilities under study but had previous experience with survey implementation. She reviewed the questionnaire with each woman and verified their understanding of the topics and situations addressed in the instrument. As part of the orientation, the nurse recommended that the mystery patients: behave naturally when interacting with health workers, not reveal to the staff or health workers that they are mystery patients, pay attention to exam facilities and the questions and information asked, and phone the research team nurse to debrief about their experiences soon after their visits finished. The nurse developed a timetable for the mystery patients and saved their phone numbers in order to remind them - with their consent - about their exam dates and responsibilities related to the research. The week of the scheduled exams, the nurse called the women to remind them about the appointment date and time. After the Pap test exams, another timetable was developed related to obtaining the test results. The nurse called the women to remind them to get their test results as scheduled and reiterated the importance of calling her to debrief about results delivery.
Data collection and analysis
The mystery patients had their exams at their local health facilities during September, October, and November 2017. Two mystery patients had their Pap test exams at the Provincial 2nd level hospital, four mystery patients at local health centers, and two at an outreach clinic set up at the local school by a non-governmental organization (NGO) providing services to the community. The research team nurse used the debriefing questionnaire to collect information from the mystery patients via telephone after their exams. The nurse also talked with the mystery patients after they received their test results to record information on the results delivery process. The data was later coded and entered into an Excel spreadsheet for quantitative data analysis and description.
Ethical approval
During study design development and application for funding, ethical concerns were raised about the mystery patient method. To avoid potential issues related to participants' ongoing health needs after Pap test exams, we decided to recruit women who already had a Pap test exam scheduled. In this way, the mystery patient debriefing questionnaire was similar to an exit interview. The local health officials did not want to announce or obtain consent from health facility managers or health care providers before the research started because they felt that if clinic managers or providers knew the mystery patients would be visiting, the results could be biased. Informed consent documents outlined the requirements and conditions for the mystery patients to participate in the study. The informed consent document also highlighted the availability of a Bolivian psychologist who would be on standby for the mystery patients to contact after their exams as needed for issues that could arise during the Pap test exam procedure. The research study protocol was approved by the Touro University California IRB (# PH-5517-TW) on February 15, 2017, and the Comité de Bioética de la Facultad de Medicina, Universidad Mayor de San Simon, Cochabamba - Bolivia on February 20, 2017.
RESULTS
Eight female mystery patients aged 28-35 assessed the quality of health facilities, health services, and the delivery of test results. Information on the mystery patient sample characteristics are outlined in Table 1.
Location (n=4) | Health Facility (n=4) | Mystery Patients (n=8) |
---|---|---|
Community A | Hospital (2nd level) | 2 |
Community B | Health Center B | 2 |
Community C | Health Center C | 2 |
Community D | NGO Outreach Facility | 2 |
Performance of the tool
See Table 2 for a summary of the tool's performance measures and Table 3 for an abridged report of patient's responses to services quality questions. The complete questionnaire and an unabridged table of responses are available on request. All women attended their screening visits and participated in follow-up debriefings. Three patients had to go to the health facility more than once to be screened. The research nurse was able to collect both the initial debriefing information and the follow-up test result information from the mystery patients using the entire questionnaire. The nurse did have difficulty tracking down some of the mystery patients after their exams, and a period of almost three weeks passed before some shared their information.
n | % | |
---|---|---|
Mean number of non-responses for items per questionnaire | 4.88 (± 2.53) | 7 |
Questionnaire items with non-responses or skip pattern problems | 15 | 22 |
Non-responses | 12 | 18 |
Skip-pattern problems | 3 | 4 |
Topic | Item | Response | n (%) |
---|---|---|---|
Health facility environment | Facility has signage for Pap test | Yes | 1 (12.5) |
Reception staff kind (very kind / kind) | Yes | 5 (62.5) | |
Respectful treatment at reception | Yes | 4 (50.0) | |
Payment required for consultation | Yes | 1 (12.5) | |
Outside | 6 (75.0) | ||
Cleaning (very clean/clean) | Inside | 7 (87.5) | |
Bathrooms | 3 (37.5) | ||
Waiting room availability | Yes | 5 (62.5) | |
Information about cervical cancer on walls | Yes | 3 (37.5) | |
Information about Pap test delivered by provider | Yes | 5 (62.5) | |
Information in the facility about reproductive/sexual rights for women | Yes | 2 (25.0) | |
Pap test availability (at least one day each week) | Yes | 2 (25.0) | |
Health services | Waiting time perception (from registration to examination) | Very long Long | 3 (37.5) 5 (62.5) |
Provider informs about consultation secrecy | No | 8 (100) | |
Pap test process explained | No | 8 (100) | |
Treatment options (for positive Pap test) explained | No | 7 (87.5) | |
Return to facility several times before having the exam | Yes | 3 (37.5) | |
Provider used gloves during Pap test | Yes | 8 (100) | |
Pap test performed in a private setting | Yes | 6 (75.0) | |
Pap test felt rushed | Yes | 7 (87.5) | |
Period of time available to ask questions about Pap results | Yes | 1 (12.5) | |
Communicational materials about Pap test delivered | No | 8 (100) | |
Provider listens carefully to questions Provider delivers clear answers to questions | No No | 8 (100) 8 (100) | |
Women felt comfortable and their dignity was respected during examination | Yes | 2 (25.0) | |
Women felt treated with respect during examination | Yes | 6 (75.0) | |
Women felt the provider was concerned about their comfort | No | 8 (100) | |
Importance of returning for Pap test results was explained | Yes | 2 (25.0) | |
Time to return to get Pap test results was informed | Yes | 7 (87.5) | |
Pap test exam consultation satisfaction | Very satisfactory/ satisfactory | 2 (25.0) | |
Pap test information satisfaction | Very satisfactory/ satisfactory | 2 (25.0) | |
Payment when collecting Pap test results | Yes | 2 (25.0) | |
Based on experience, women would recommend Pap test to friends | Yes | 8 (100) | |
Pap test results | Results delivery performed in indicated time | Yes | 2 (25.0) |
Results meaning explained by provider | No | 7 (87.5) | |
Completely understood results | No | 6 (75.0) | |
Provider asked if women had questions about the results | No | 7 (87.5) | |
Women had questions about the results | Yes | 5 (62.5) | |
Results were given to someone else | Yes | 2 (25.0) | |
Need for further treatment was informed | Yes | 2 (25.0) | |
Informed about consequences of not having treatment | Yes | 2 (25.0) | |
Treatment was delivered | Yes | 1 (12.5) | |
Informed about where and how to get treatment | Yes | 1 (12.5) | |
Information about when to return for another Pap test was delivered | Yes | 5 (62.5) | |
Sexual and reproductive health information was delivered | No | 8 (100) |
*Note: An unabridged table of questionnaire responses is available upon request.
On average, 93% of questions in the questionnaire were answered completely. The number of non-responses per questionnaire ranged from one to nine, with an average of 4.88 (standard deviation = 2.53). These non-responses and problems with skip patterns were located in 15 of the 68 questionnaire items (22%). They were concentrated in the questions about the health facility environment and the behavior of non-clinical staff. Within this cate-gory, the questionnaire items with the highest number of non-responses asked about fees receptionists' attitudes. Non-responses were highest in the two patients whose screening took place at an ad hoc NGO screening center located in a school building and who were, therefore, unable to answer questions about particular physical aspects of the health facility (e.g., posters on waiting room walls).
Other than this issue, which had not been flagged in initial cognitive interviews with local women, the research team nurse and mystery patients had little difficulty understanding or answering debriefing questions. While there were a few questions with violated skip patterns, there were no other instances of consistently in-complete or inappropriate responses.
Ethical issues
There were no reported adverse consequences during the exams, and none of the mystery patients requested to speak with the psychologist who was available after their exams to discuss concerns. None of the health facility managers or health workers reported knowing anything about the mystery patient visits. All eight of the mystery patients reported their findings after their exams without any mention of adverse advents.
Health facility environment
The mystery patients reported that only the hospital offered Pap exams at least once a week. The health centers did not offer weekly Pap exams. Only the hospital had cervical cancer posters on the walls and information about wo-men's reproductive health rights. The health centers did not have this information available to patients (Table 3).
Quality of Services
The mystery patients reported that 100% of the health facility staff and providers were friendly. All (100%) of the health workers at the facilities used gloves during the Pap exams. In 75% of the health facilities, exam rooms were in a private location where others could not hear conversations or see examinations.
In none of the health facilities were the mystery patients told what would happen during the exam or that the results would be kept confidential, or what treatment options they had if the Pap test results were abnormal. The mystery patients reported that only one health worker as-ked the mystery patient if she had any questions about the exam. None of the health workers encouraged the mystery patients to ask questions. Only at the NGO outreach clinic did the mystery patients feel comfortable with their dignity respected during the exam. Only 25% of the health workers explained the importance to the mystery patients of following up to get their test results (Table 3).
Pap test results delivery
The mystery patients reported that only hospital health workers gave them test results in the time period indicated. Only mystery patients visiting the hospital completely understood their Pap test results; the other six mystery patients did not (Table 3).
DISCUSSION
This pilot study assessed the feasibility of the mystery patient method for evaluating the quality of cervical cancer screening. The method effectively collected detailed information on Bolivia's cervical care screening program based on national and international standards. The mystery patient method was successfully implemented to collect information related to cervical cancer screening services and practices. Questionnaires were on average 93% complete, with non-responses largely stemming from differences in screening settings which rendered some questions irrelevant. There were no reported adverse events from under-going the screening or completing debriefing and no complaints from health facility managers or health workers.
While our study is small, it is important to note that almost no other studies in the past decade report on the quality of Bolivia's existing Pap test exam screening strategy rather than new strategies such as HPV self-testing. Our findings suggest that cervical cancer screenings in this rural community operate sporadically, with key quality components lacking. Health workers are friendly but do not encourage patients to ask questions or communicate necessary information to them. The results suggest that there remain considerable challenges related to cervical screening follow-up and obtaining laboratory results.
The Bolivia MoH has focused on increasing the numbers and quality of human resources for health in rural areas and establishing digital health interventions such as telemedicine to bring services to communities. These initiatives have not been implemented to scale or evaluated 31,32. If the reports of our mystery patients are representative of other rural communities in Bolivia, they suggest that these interventions and other access and quality-improvement strategies must be re-vised and expanded, with a focus on strengthening the health system. Such further refining and targeting of improvements will require ongoing monitoring, evaluation, and research. This study has piloted a method and tool that may help in these efforts and has demonstrated that it can be employed without adverse impacts on patients or providers.
Several factors were important for the successful fielding of this method. First, early in the research design phase, the research team, which included male and female, clinical and public health, researchers from Bolivia and the USA, discussed and addressed potential ethical issues. Second, working with the local Bolivian health authorities to design the research was an effective and necessary process. Third, the researchers recruited motivated mystery patients due to the time and effort spent working with health and rural social authorities. Importantly, we were able to recruit women already scheduled for cervical cancer screening. Because of the researchers' preparation and high level of professionalism, they were welcomed in the field.
This study is the first to our knowledge to use the mystery patient method to assess the quality of cervical cancer screening in a LMIC setting. With fine-tuning, it can be scaled up and effectively implemented for larger samples to collect data needed to improve cervical cancer screening and elimination strategies. Evaluating reproductive health services from a mystery patients' viewpoint will provide new perspectives for more comprehensive evaluations over time.
Suggested improvements of the tool and method based on our experience include adding skip patterns or "non-applicable" responses for questions about health facility infrastructure that might not be relevant to screenings at outreach events or other locations. Questionnaire items may also need to be added or removed to address different types of screening offered at facilities (e.g., HPV testing or VIA). Data collectors will require rigorous training to ensure that questionnaires are completed fully. We would also suggest electronic rather than paper data collection/debriefing tools in the future to assist in flagging missing questions, skip pattern violations, and respondents in need of follow-up.
Limitations
There were limitations in the design and implementation of this pilot study. The small sample size meant that the sample was not representative of the community. Recall bias was a concern for mystery patients contacted only after several attempts. In addition, although patients were free to voice questions and concerns during debriefing sessions and were asked how they felt about the exam itself (Table 3), we did not systematically collect information on how patients felt about the debriefing process or how comfortable they felt in their mystery patient role. Future research should address this gap. Finally, we could not assess the quality of Pap test interpretations by pathologists using the mystery client method.
This study suggests that the mystery patient evaluation method may be a valuable tool in scaling up cervical cancer screening and elimination strategies in Bolivia and elsewhere in the region. This method could generate insightful, useful, and actionable information and facilitate health rights education by providing reliable first-hand evidence to communities and health authorities. When communities know their health care rights and demand quality reproductive services, health care quality will improve in rural Bolivia ♠