TY - JOUR
AU - Krezanoski, Paul, J
AB - Abstract Research on health systems in resource-limited settings has garnered considerable attention, but the dispensing of individual prescriptions has not been thoroughly explored as a specific bottleneck to effective delivery of care. The rise of human immunodeficiency virus/tuberculosis prevalence and non-communicable diseases in the Kingdom of eSwatini has introduced significant pressures on health facilities to meet patient demands for lifelong medications. Because automated pill counting methods are impracticable and expensive, most prescriptions are made by means of manually counting individual prescriptions using a plastic dish and spatula. The aim of this work was to examine the perceptions of health providers of causes for pill counting errors, and pill counting’s impact on clinic workflow. Our study took place in 13 randomly selected public health facilities in eSwatini, stratified by three groups based on monthly patient volumes. Thirty-one participants who count pills regularly and 13 clinic supervisors were interviewed with semi-structured materials and were audio-recorded for later transcription. Interviews were thematically analysed with inductive coding and three major themes emerged: workflow, counting error causes and effect on clinic function. Findings demonstrate large variety in how facilities manage pill counting for prescription making. Due to patient demands, most facilities utilize all available personnel, from cleaners to nurses, to partake in prescription making. Major causes for pill counting errors were distractions, exhaustion and being hurried. Participants mentioned that patients said that they had initially received the wrong quantity of pills and this affected medication adherence measurements based off pill counts. Most participants described how efforts put into pill counting detracted from their work performance, wasted valuable time and increased patient wait times. Future research is needed to quantify prescription accuracy, but our data suggest that interventions directly alleviating the burden of pill counting could lead to improved clinic quality and possibly improve patient outcomes. Clinic quality, pharmaceutical delivery, medication adherence, pill counting Key Messages Repackaging medications from bulk delivery to individual prescriptions may detract from the overall quality of care that health facilities are otherwise potentially able to deliver. Pill counting affects clinic quality through inaccurate pill counts, and causes for pill counting errors were identified as distractions, lack of mental energy and being hurried. Pill counting also affects clinic quality through interference with health facility work flow, and clinical staff perceive that efforts put into pill counting can be deleterious to the quality of their patient care. Inaccurate initial pill quantities can detract from the usefulness of medication adherence monitoring and health education that relies on pill counts to assess adherence. Introduction Enhancing the efficiency and quality of health care in low-resource settings by 2030 is a key area of interest for the World Health Organizations (WHO) Sustainable Development Goals (Kieny et al., 2017). This is informed by both the increase in the prevalence of non-communicable diseases such as hypertension and diabetes and the increase in the prevalence of communicable diseases requiring chronic treatments such as human immunodeficiency virus (HIV)/acquired immune deficiency syndrome (AIDS), which together place tremendous strain on historically fragile health systems and facilities. One of the specific goals set by the United Nations (2012) in the 2030 Agenda is to ‘… provide access to affordable essential medicines’, and yet medication procurement and distribution is inefficient in many regions of the world (Yip and Hafez, 2015). The Kingdom of eSwatini (formerly Swaziland) is a nation where improving the efficiency of pharmaceutical access and delivery to patients would prove to be particularly beneficial. This sub-Saharan African country of 1.1 million people has the world’s highest prevalence of HIV (27.3%) (The World Bank, 2017a,b). Despite 63% of the population living below the poverty line and 76% residing in rural areas, 79% of people living with HIV are on treatment, which is 19% higher than the average for 21 countries in Eastern and Southern Africa [Joint United Nations Programme on HIV/AIDS (UNAIDS), 2017; The World Bank, 2017c,d]. Achieving these high HIV treatment rates is made more challenging by low numbers of trained medical providers such as doctors, nurses and midwives. The eSwatini Ministry of Health’s (2012) latest report showed that there were 1.69 medical providers per 1000 people which is far below the 4.45 threshold in WHO’s Sustainable Development Goals (World Health Organization, 2016). In eSwatini, nurses often take on the primary provider role, such as diagnosing, prescribing and creating treatment plans (Ugochukwu et al., 2013). Although there are plans for increasing the number of pharmacy technicians, latest eSwatini Ministry of Health (2013) reports show that there are only 20 pharmacists and 44 pharmacy technicians in country. This lack of personnel necessitates that many pharmaceutical dispensaries must rely on nurses, health aides and even non-health professional staff such as cleaners and security guards whom may have limited training, to count pills, package prescriptions, dispense and provide pharmacy education to patients. Most medications in eSwatini are shipped to patient care facilities from a central store in large bottles with quantities ranging from 500 to 5000 pills. Compared with medication blister packs or smaller bottles (<100 pills each), bulk delivery keeps costs down by reducing packaging, shipping and storage costs (Pilchik, 2000). However, these bulk bottles require recounting before provision to patients. In higher income countries, monthly pill quantity prescriptions (e.g. 30, 60, 120) are typically counted with electronic counting machines because this requires less effort, saves time and is typically more accurate than counting pills by hand (Fung et al., 2009). In low-resource settings such as eSwatini, however, electronic pill counting machines are rare due to their cost and requirements for maintenance, unreliable electricity, humid/hot environmental conditions and security concerns over housing an expensive machine. As a result, the most common method of preparing a prescription in eSwatini, in accordance with WHO policies, is to pour pills from the large bottles onto a plastic dish, count them individually using a plastic or wooden spatula and then package in plastic bags for provision to patients (Management Sciences for Health, 2012). Improving the quality of the provision of medications in low-resource environments requires an understanding of the barriers to efficient care in pharmaceutical dispensing. We undertook this qualitative study to explore the perceptions of both pill counters and clinic supervisors related to pill counting processes, causes and impacts of pill counting errors and overall views of how pill counting impacts clinic quality. Materials and methods We used interviews and structured surveys to examine the perceptions of two distinct populations in relation to pill counting: (1) health facility supervisors and (2) pill counting personnel. Each population was asked the same eight semi-structured interview questions, but each group was orally administered a different structured survey. Health facility supervisors were asked 15 health facility characteristic questions and pill counting personnel were asked 13 demographic/counting characteristic questions. Data collected from the structured surveys were tallied and then tabulated. This qualitative study was performed concurrently with a second study, which will be published separately and which involved testing an alternative pill counting method. During the same site visits, only the pill counting personnel were taken aside before the interviews to test a non-electric pill counting solution. After testing the device, a brief 12-question Likert scale survey was administered to measure device acceptability. Every effort was taken to preclude the likelihood of the tandem study shaping participant’s viewpoint of pill counting activities. This included, but was not limited to, training research assistants (RAs) to reassure participants that all data would be anonymously collected and analysed and that there were no right or wrong answers as the results of the study had no bearing on the future of their work. The Ministry of Health provided a comprehensive list of all 207 public health facilities in eSwatini. We screened out specialty clinics [e.g. prisons, tuberculosis (TB) clinics, mental health centres] and facilities that served <500 patients per month, leaving a sample of 124 facilities. We stratified these 124 facilities into three groups based on monthly patient volume: small (500–1100 patients/month), medium (1101–2200 patients/month) and large (>2201 patients/month). We randomly selected 11 facilities from each stratum and then chose 5 facilities from each stratum while trying to balance representation from eSwatini’s four regions. Our selection included six facilities from Hhohho, four facilities from Manzini, three facilities from Lubombo and two facilities from Shiselweni. During the study, 2 small sized facilities of the initially selected 15 facilities were not used due to transportation logistics. Facility replacements were not feasible for either clinic due to time constraints (Table 1). Table 1 Clinic characteristics (n = 13) Region, n (%)  Hhohho 6 (46.2)  Manzini 4 (30.8)  Lubombo 2 (15.4)  Shiselweni 1 (7.7) Size and patients per month, n (%)  Small (500–1100) 3 (23.1)  Medium (1100–2200) 5 (38.5)  Large (>2200) 5 (38.5) Clinic staff (mean ± SD) 37.3 ± 40.2 Range: 7–149 Staff engaged counting per day (mean ± SD) 5.9 ± 6.5 Range: 2–27 Region, n (%)  Hhohho 6 (46.2)  Manzini 4 (30.8)  Lubombo 2 (15.4)  Shiselweni 1 (7.7) Size and patients per month, n (%)  Small (500–1100) 3 (23.1)  Medium (1100–2200) 5 (38.5)  Large (>2200) 5 (38.5) Clinic staff (mean ± SD) 37.3 ± 40.2 Range: 7–149 Staff engaged counting per day (mean ± SD) 5.9 ± 6.5 Range: 2–27 SD, standard deviation. Open in new tab Table 1 Clinic characteristics (n = 13) Region, n (%)  Hhohho 6 (46.2)  Manzini 4 (30.8)  Lubombo 2 (15.4)  Shiselweni 1 (7.7) Size and patients per month, n (%)  Small (500–1100) 3 (23.1)  Medium (1100–2200) 5 (38.5)  Large (>2200) 5 (38.5) Clinic staff (mean ± SD) 37.3 ± 40.2 Range: 7–149 Staff engaged counting per day (mean ± SD) 5.9 ± 6.5 Range: 2–27 Region, n (%)  Hhohho 6 (46.2)  Manzini 4 (30.8)  Lubombo 2 (15.4)  Shiselweni 1 (7.7) Size and patients per month, n (%)  Small (500–1100) 3 (23.1)  Medium (1100–2200) 5 (38.5)  Large (>2200) 5 (38.5) Clinic staff (mean ± SD) 37.3 ± 40.2 Range: 7–149 Staff engaged counting per day (mean ± SD) 5.9 ± 6.5 Range: 2–27 SD, standard deviation. Open in new tab The study took place between July and August 2016. All facilities were notified that our research staff would be visiting to conduct interviews prior to their arrival. Pill counters were identified by supervisors as personnel who were at least 18 years and had performed pill counting by plastic dish and spatula at least twice within the last month. Only pill counters present at the facility on the day of our visit were included. In a private room, RAs conducted an ∼30-min, one-on-one, audio-recorded interview with participants at the end of the planned pill counting activities for the other study (see above). RAs informed potential participants of the purpose of the study and obtained written informed consent. Data were collected on demographics, personal pill counting activities (e.g. hours per day), clinic roles, perceptions of pill counting and the effect of pill counting on other clinic duties. All interviews were conducted in English or siSwati according to the participant’s preference. The facility supervisors were identified upon arrival at the facility as the most senior staff member present who physically worked within the facility grounds. In the supervisor’s private room, an RA conducted a 30–60-min, one-on-one, audio-recorded interview with the supervisor. Data were collected with a semi-structured interview regarding the supervisor’s perceptions of pill counting and its impact on their facility. All interviews were conducted in English. Supervisor and pill counter data were inputted and later analysed using a combination of Microsoft Excel and Stata (version 15). Semi-structured audio-recorded interviews were transcribed verbatim and, as necessary, translated into English by an RA fluent in both languages. Audio recordings of three pill counters and two supervisor interviews were lost due to data corruption. Pill counters and supervisor interview transcripts were inductively coded (bottom-up approach) by JDK and PK using the thematic analysis tool Dedoose (version 8.0.42). We used Braun and Clarke’s (2006) six-phase qualitative analysis as a general guideline to iteratively process our interview data. Phases included ‘familiarize yourself with the data’, ‘generating initial codes’, ‘searching for themes’, ‘reviewing themes’, ‘defining and naming themes’ and ‘producing the report’. Supervisor and pill counter interviews were coded separately, but themes included both populations since there was considerable overlap between the two. Both JDK and PK identified representative quotes for the subthemes. Per Braun and Clarke (2006), we derived themes organically without the use of any pre-existing theories or frameworks. While a certain level of interpretation of the data will always exist while themes emerge, we focused efforts on using this bottom-up approach to organize transcript data from what participants actually said, rather than a deeper or abstract interpretations of the data. These themes and subthemes were processed and organized through multiple iterations until they were both ‘sufficiently distinct and internally homogenous’ (Braun and Clarke, 2006). This study received Institutional Review Board approval and Scientific and Ethics Committee approval from the authors’ institutes. Written informed consent was obtained from all participants prior to any data collection. Results Three major themes, each with subthemes, were identified in our interviews about pill counting: workflow, counting error causes and effect on clinic function (Figure 1). Figure 1 Open in new tabDownload slide A illustrative roadmap of the themes and subthemes. Figure 1 Open in new tabDownload slide A illustrative roadmap of the themes and subthemes. Workflow While the workflow of pill counting between sites varied greatly, there was overlap. Participants mostly described three pill counting activities: (1) counting during medication adherence checks, (2) counting as needed and (3) counting during pre-packaging for future patient demand. Workflow: Adherence checks Pill counting during medication adherence checks involved comparing expected pill counts, based on the last prescription, to the number of pills remaining in the patient’s possession and recording a percentage of adherence. This was noted for medications such as antiretroviral medications for HIV, opportunistic infection prophylaxis such as sulfamethoxazole/trimethoprim [cotrimoxazole (CTX)], TB medications and anti-hypertensive drugs. An HIV counsellor described what happens after an adherence check indicates that the patient may not be taking their medication properly: ‘[With our calculations] we know how much [medication] we gave them. When they return, we do the pill counting and then see how they were taking the pills. If they did not take them properly, we cut down [the pill quantity of their next prescription]. [For example], if we gave them 2 month of stock, we will go down to one month [for their next prescription]. If they still can’t take them properly … we would give them 15 days [of medication]’ (Participant 31, female, HIV counsellor). Workflow: Counting as needed Pills are counted as needed whenever prescriptions are needed but are not routinely pre-packed medications, such as those dispensed in low volumes, or when supplies of routinely pre-packaged prescriptions are exhausted. This type of counting typically occurs immediately when the patient arrives to the pharmacy to pick up their medications. One nurse explained how in their heavily understaffed facility that immediately after deciding what prescription their patient needed, they would count the pills themselves and make the prescription: ‘After seeing a patient in the consultation room and prescribing them medication, you then have to give them what you have prescribed. So, you have to count the pills [that you just prescribed] and give to the patient so they can go home’ (Participant 14, male, nurse). Workflow: Pre-packaging To save time, commonly dispensed prescriptions were often counted and packaged prior to being prescribed for a patient. In all sites visited, this accounted for the greatest proportion of time and effort for pill counting activities. Regular pill counters often perform this pre-packaging of pills as their main role through most of the day. Other staff report participation in this type of counting after their other roles are completed: ‘My job is to clean, cleaning the clinic …. I scrub, clean windows, and pick some paper waste that are being disposed by people. After finishing that I then move on to pill counting and do it until knock off time’ (Participant 16, male, cleaner). Many participants also noted counting sporadically over the course of the day as time permitted. The frequency of counting per person varied considerably in a week depending upon patient volume and drug availability, ranging from daily to once weekly. Participants mostly described how their facility would combine a scheduled morning or late afternoon counting session with additional pre-packaging when time was available. Once assembled, pre-packaged prescriptions are sorted into bins designated for that specific medication dose and quantity for provision to patients. Participants reported that CTX pills were pre-packaged in higher volumes than any other drug. Overall, the average number of CTX bottles pre-packaged per facility was 17.5 bottles per week (17 500 pills). Workflow: Counting personnel The personnel responsible for counting pills vary by location. In some facilities, pill counting rotated between people on a weekly basis, and in other sites, it was expected that pre-packaging was a facility-wide task for all staff. The vast majority of participants (80.3%) reported being either an HIV counsellor, cleaner or nurse (Table 2). A few staff who primarily performed cleaning activities reported acting as pseudo-pharmacists, counting and dispensing medication to patients without any formal training. As one cleaner explained his reluctant role in the pharmacy: ‘… it is not a job I went to school for or that I should be doing … you get scared when the person is asking you about the pill … [because] you don’t know exactly how the pill helps the person’ (Participant 4, male, cleaner). Even when facilities do employ pharmacy technicians, counting and packaging prescriptions still requires delegation to other staff members. As one supervisor talked about her pharmacy technician: ‘… he is alone there, with the cleaners who are helping him with packaging the tablets … He is really worn out because … he cannot take a day off, except on Sundays …’ (Supervisor 4, female). Table 2 Participant characteristics (n = 31) Age (mean ± SD) 40.0 ± 7.7 Range: 25–55 Gender (female), n (%) 22 (71) Clinic role, n (%)  HIV counsellor 11 (35.5)  Cleaner 10 (32.3)  Nurse 4 (12.9)  Pharmacy technician 2 (6.5)  Others (e.g. TB counsellors, health motivator) 4 (12.9) Number of years in school, n (%)  No school 2 (6.5)  Primary school 2 (6.5)  Some high school 12 (38.7)  Completed high school 9 (29.0)  Some university 1 (3.2)  Completed university 6 (16.1) Age (mean ± SD) 40.0 ± 7.7 Range: 25–55 Gender (female), n (%) 22 (71) Clinic role, n (%)  HIV counsellor 11 (35.5)  Cleaner 10 (32.3)  Nurse 4 (12.9)  Pharmacy technician 2 (6.5)  Others (e.g. TB counsellors, health motivator) 4 (12.9) Number of years in school, n (%)  No school 2 (6.5)  Primary school 2 (6.5)  Some high school 12 (38.7)  Completed high school 9 (29.0)  Some university 1 (3.2)  Completed university 6 (16.1) SD, standard deviation. Open in new tab Table 2 Participant characteristics (n = 31) Age (mean ± SD) 40.0 ± 7.7 Range: 25–55 Gender (female), n (%) 22 (71) Clinic role, n (%)  HIV counsellor 11 (35.5)  Cleaner 10 (32.3)  Nurse 4 (12.9)  Pharmacy technician 2 (6.5)  Others (e.g. TB counsellors, health motivator) 4 (12.9) Number of years in school, n (%)  No school 2 (6.5)  Primary school 2 (6.5)  Some high school 12 (38.7)  Completed high school 9 (29.0)  Some university 1 (3.2)  Completed university 6 (16.1) Age (mean ± SD) 40.0 ± 7.7 Range: 25–55 Gender (female), n (%) 22 (71) Clinic role, n (%)  HIV counsellor 11 (35.5)  Cleaner 10 (32.3)  Nurse 4 (12.9)  Pharmacy technician 2 (6.5)  Others (e.g. TB counsellors, health motivator) 4 (12.9) Number of years in school, n (%)  No school 2 (6.5)  Primary school 2 (6.5)  Some high school 12 (38.7)  Completed high school 9 (29.0)  Some university 1 (3.2)  Completed university 6 (16.1) SD, standard deviation. Open in new tab Counting error causes Participants discussed three potential causes for pill quantity inaccuracies. Counting error causes: Distractions Distractions were commonly mentioned as a cause of pill counting inaccuracies. Of the distraction types, cell phone ringing, talking and unspecified other noise were most frequently discussed. Cell phones were ubiquitous among staff personnel and many described how cell phone rings could interrupt their counting. The constant flow of people going in and out of the room was another form of distraction. None of the facilities we visited had a designated quiet pill counting room. Instead, the location of pill counting was determined by the availability of space: ‘… You see that this [dispensary] is one room? The medication is dispensed to the patients right here …. and then one of [the patients] are shouting … that disturbs us’ (Participant 28, female, cleaner). Many of the participants described situations where they were constantly surrounded by others, either patients or other staff personnel: ‘… you sometimes find that whilst you’re counting you forget … there’s an interesting story that someone shares and it makes you laugh, and you then lose concentration and forget how many [pills] there are, and … I then just guess what number I remember saying and you find that I have added an extra 5 …’ (Participant 21, male, HIV counsellor). Nurses particularly discussed that attending to emergencies was a major cause of errors: ‘You find that an emergency occurs and you have to leave and attend to it. In that rush, it’s possible to unknowingly close a prescription bag without putting [enough] pills inside …’ (Participant 14, male, nurse). Some described times where other staff members interrupted them, without consideration for the task they were engaged in. Some participants expressed frustration with these distractions, and it required them to find a private area to count to reduce distractions: ‘I have to find a corner in which to hide myself and count until I finish and then take the pills to the dispensary’ (Participant 23, female, HIV counsellor). Counting error causes: Exhaustion Many participants cited an inability to focus because pill counting was an exhausting task: ‘… you were very busy [throughout the day] and now you have to count pills, your mind is exhausted’ (Participant 3, female, other). For some participants, emotional and compassion fatigue factored as a major reason for counting errors. HIV counsellors, TB counsellors and nurses, all of whom engage with patients on an intimate level, described how draining it was managing patients living in poverty with chronic illnesses like HIV/AIDS. One HIV counsellor, who frequently does medication adherence checking, described an emotional situation that directly disrupted her ability to focus on pill counting: ‘[A mistake] … happened while I was pill counting for the adherence of a client who was struggling accepting [her HIV status]. Her situation emotionally touched me …. She said she was having it hard, her husband threw her pills away, she was dependent on the very husband, had no background, no parents. So … I got distracted and [felt] drained, yet I was supposed to count pills and I would find myself counting 40’s yet I was [actually] in the 30’s … You find that the adherence session is draining’ (Participant 19, female, HIV counsellor). Counting error causes: Hurried Twenty-eight participants said that a major cause of errors was feeling rushed because patients were waiting or they had other work to attend to. Furthermore, survey results showed that 61.3% of pill counters spent at least 40% of their efforts at work counting pills (Table 3). Nearly all personnel shared a sense of urgency to attend to tasks other than counting pills and felt that this contributed to haphazard counting: ‘You find you have 10 000 pills that have to be counted in one day, not even the entire day but at a particular time. Then they give you maybe 5 containers of [a diabetes medication] which also contain 1000 tablets … immediately after counting, they give you calcium to count … and you may find that by the time it is on the new pill, you have already made a blunder’ (Participant 29, female, cleaner). Many participants explained that they had to adjust their counting quality according to patient demand, noting that this was the only way for them to give medications to patients in a timely manner. Participants described that patients often have to wait for long periods of time in line, creating a source of pressure: ‘I do not know whether I have counted the right number or not. I cannot afford to empty what I have already counted … but {deep breath}, I think of the fact that I have limited time, and there is this long queue outside’ (Participant 6, female, HIV counsellor). Others described a situation where they saw a co-worker make an error, but due to the difficulty of the job, they did not correct their error. ‘It happens that someone counted pills and they were not enough, another you will hear saying, ‘oh shame, I forgot how many they are’, but continue to pack … and doesn’t start afresh … they tell themselves that the number they have in mind is the one they counted, whereas it is not the case. Ok, mistakes do happen in this manner because, hey!, it is tough’ (Participant 25, female, HIV counsellor). Another described a situation where they knowingly gave the wrong prescription quantity to patients because of the necessity to count quickly to give patients their medications: ‘You are worried that [the patients] have been waiting for a long time, so you end up putting pills into bags without counting them … Question: You would not be counting at all? Participant: Yes’ (participant 6, female, HIV counsellor). Table 3 Participant counting characteristics (n = 31) Estimated percent effort counting pills (%), n (%)  0–20 7 (22.6)  20–40 5 (16.1)  40–60 10 (32.3)  60–80 4 (12.9)  80–100 5 (16.1) Average hours per day counting pills, n (%)  0–2 13 (41.9)  3–5 10 (32.3)  6–8 7 (22.6)  >9 1 (3.2) Average days per week counting pills, n (%)  0–1 4 (12.9)  2–3 6 (19.4)  4–5 15 (48.4)  6–7 6 (19.4) Estimated percent effort counting pills (%), n (%)  0–20 7 (22.6)  20–40 5 (16.1)  40–60 10 (32.3)  60–80 4 (12.9)  80–100 5 (16.1) Average hours per day counting pills, n (%)  0–2 13 (41.9)  3–5 10 (32.3)  6–8 7 (22.6)  >9 1 (3.2) Average days per week counting pills, n (%)  0–1 4 (12.9)  2–3 6 (19.4)  4–5 15 (48.4)  6–7 6 (19.4) Open in new tab Table 3 Participant counting characteristics (n = 31) Estimated percent effort counting pills (%), n (%)  0–20 7 (22.6)  20–40 5 (16.1)  40–60 10 (32.3)  60–80 4 (12.9)  80–100 5 (16.1) Average hours per day counting pills, n (%)  0–2 13 (41.9)  3–5 10 (32.3)  6–8 7 (22.6)  >9 1 (3.2) Average days per week counting pills, n (%)  0–1 4 (12.9)  2–3 6 (19.4)  4–5 15 (48.4)  6–7 6 (19.4) Estimated percent effort counting pills (%), n (%)  0–20 7 (22.6)  20–40 5 (16.1)  40–60 10 (32.3)  60–80 4 (12.9)  80–100 5 (16.1) Average hours per day counting pills, n (%)  0–2 13 (41.9)  3–5 10 (32.3)  6–8 7 (22.6)  >9 1 (3.2) Average days per week counting pills, n (%)  0–1 4 (12.9)  2–3 6 (19.4)  4–5 15 (48.4)  6–7 6 (19.4) Open in new tab Effect on clinic function Participants were asked about medication adherence checks and if and how pill counting interferes with their clinical role and how their work would be impacted if they no longer had to perform this task. During the interviews, participants described situations where patients said they received the wrong number of medicine pills, how pill counting activities was related to increasing wait times and how pill counting impacted their quality of work, was perceived as wasted time and overall decreased the quality of care provided to patients. Effect on clinic function: Medication adherence monitoring We asked participants how medication adherence checks are generally conducted and which medications are routinely checked. In all of our visits, antiretroviral (ART) medications were always monitored for adherence, while CTX medications were checked for adherence by nearly all of the health facilities. A few of the health facilities also did adherence checks for other drugs such as hypertension medication. Many mentioned that patient adherence was based on a pill count of the pills remaining in the patient’s possession at the time of the visit. If a patient had more pills than expected (not enough pills were taken) or fewer than expected (excessive pills were taken), they would be considered non-compliant. Some medications are initially given in a 1-week supply, and pending the patient’s adherence, they can eventually receive a 3-month supply, greatly improving convenience. When asked for potential explanations for non-adherence, a quarter of the participants cited inaccurate initial pill counts. One nurse described such a situation in which a patient who had been labelled non-adherent argued that they received the wrong number of pills to begin with: ‘…. the patient may say, ‘it means you gave me a lot of pills’. When I redo the pill count again … for instance for CTX, and I tell them that they didn’t take medication correctly, they restate saying ‘no, it means you gave me more [extra] pills.’ They say that they take the medication, and highlight that they wouldn’t get the pills at the clinic without an intention to take them’ (Participant 11, female, HIV counsellor). Another nurse described a similar situation. A patient returning for a medication adherence check, who was supposed to have received a 3-month supply of medication, had instead claimed to have run out of medication earlier. Upon re-calculation, it was determined that the patient had only been given a 2-month supply. ‘… we had supplied [the client] with three months of medication … [but] we have CTX for two months, and the client came back to report that pills had run out along the way’ (Participant 3, female, other). Some participants said that they felt sceptical of the patients’ claims and placed confidence in the adherence check over the patient’s recollections. ‘Question: have you ever had a patient argue [when they are shown to be non-compliant]? Supervisor: Yes, we usually have that but we just bring the calendar, like you see the month has 30 days, just show him this container. That’s containing 30 tablets, it was supposed to be consumed during this month … where did you get the other [extra] medication? He would just say “whoa!” Because the arguing is one way of defending themselves, but they know … what happened’ (Supervisor 5, female). Effect on clinic function: Decreased job performance Most participants said that if they did not have to count pills, their ability to perform their primary role would be improved and they would be able to fulfil these responsibilities more efficiently. A pharmacy technician explained that: ‘I’d put more time in [medicine] stock management, which would allow me to have more time [to spend] on the supply chain. There would be less drug shortages because I would have enough time’ (Participant 17, male, pharmacy technician). This response exemplifies a common thread in participant responses—namely that if pill counting was no longer assigned to them, personnel would be better able to perform the job that they were originally hired for. Regardless of the participant’s job title, there was concern about spending so much effort counting pills: ‘I think I would do the job I was hired for 100%, unlike now where I’m only doing it 49%. It would definitely make it so I can do … the job I was hired to do, and do it properly … You find yourself having to focus a lot on packing pills, and are unable to do other duties, so your job gets neglected. Most of our time is spent packing pills’ (Participant 4, male, cleaner). Effect on clinic function: Time wasting HIV counsellors, whose primary task is patient education and medication adherence, felt pressured to spend their available time counting pills, instead of educating patients. Many participants described fulfilling many roles in the clinic, all of whom suffered when they had to count out prescriptions as well. Further complicating the matter, nurses are highly valued in eSwatini, likely because nursing education represents the highest level of healthcare provider training available in the country. Accordingly, the nurses who we interviewed seemed to feel particularly frustrated by the underutilization of their training: ‘[If I didn’t have to count pills anymore], I think a lot of time would be saved. The time I find myself sitting and counting pills … You find that you waste a lot of time. For instance, tomorrow is time for me to write a report. Just imagine the time required in report writing combined with that of pill counting …. it really consumes time’ (Participant 15, male, nurse). All supervisors, also nurses by training, mentioned that their staff were overworked, their facility was understaffed and pill counting placed addition workload on their staff: ‘Yes, [we are] short staffed, so we do not have time to pack …. we don’t have a specified person in the pharmacy. We are the nurses, but we are also the pharmaceutical person’ (Supervisor 9, female). They also wished that these more highly trained individuals could spend more time on patient care and less on mundane tasks including counting prescriptions: ‘… [Nurses] should focus on just nursing, not counting and dispensing medication’ (Supervisor 2, female). And another supervisor adamantly said: ‘A nurse should be a nurse, they should not be every position. They should not be thinking of pills’ (Supervisor 10, female). Effect on clinic function: Patient wait times In most of the facilities we visited, after patients saw the nurse, they were required to wait in a line to receive their medications. If medicines were not readily available at the pharmacy, staff had to count pills as the patient waited. In our interviews, participants often cited this scenario as a significant cause of lengthening wait times: ‘[Pill counting] affects my work a lot in that you find that we run out of pills before the clients arrive, and the clients have to wait for a long period of time … while we are counting the pills’ (Participant 8, female, nurse). Another personnel member mentioned a similar experience when pre-packaged pills had run out and described the impact this had upon patients. ‘Okay, counting the pills, you find that … maybe the patient does not understand when you tell them that ‘there are no pills at this time, we still have to count the pills’ … They end up spending longer periods in the clinic. Question: How much does that interfere with your daily work? Participant: Intensely, intensely because you find that although the patient came …. at 8 am … I ended up seeing her at 2 pm. She is delayed and I would then feel bad …’ (Participant 6, female, HIV counsellor). The majority of eSwatini’s population utilize public transportation, which varies by region in terms of reliability. In one interview, the participant described how the lack of transportation contributed to time constraints by forcing patients to leave the clinic by a certain time to make their last transport option home. They explained that any extra time that the patients have to spend at the clinic increases their odds of missing transport home. Some supervisors also noted the challenge unreliable transportation posed to caring for patients. One supervisor explained how some patients spend an average of 2-h travelling either one way to or from the clinic, and another supervisor explained that transportation could take 3–4 h one way. Another explained: ‘We have to race [and count pills] … because they will miss the transport. We have to make sure they get back home on time’ (Supervisor 9, female). Effect on clinic function: Decreased quality of patient care Participants often expressed that the quality of patient care would improve substantially if they were no longer required to count pills and could increase the time for patient–healthcare worker interactions. As one nurse explained: ‘You find that instead of fully attending to a patient, I have to concentrate on pill counting. …. resulting in failure to give adequate care’ (Participant 14, male, nurse). HIV counsellors felt that their ability to talk to patients was hindered because of the excessive time needed for counting pills at a later point in the day: ‘Participant: In a case where I would no longer [be] tasked with counting pills … I would talk to [the patients] the way I’m supposed to … There would be enough [time] ’ (Participant 6, female, HIV counsellor). Healthcare workers described how the anticipatory workload of pill counting could cause them, although reluctantly, to ‘cut-corners’ on the care given to patient they were seeing in that moment: ‘ [If I didn’t have to count pills anymore], I think that I would be able to give the patient enough time for clear understanding, because sometimes you find yourself under pressure to finish the patient fast so that you are able to count pills for the following day, you see? So, you find that indeed the patient is not given the necessary attention’ (Participant 9, female, nurse). Discussion To our knowledge, this is the first qualitative study exploring the impact that pill counting has on health facilities in low-resource settings. We found substantial potential downstream impacts on patient care, quality of care and clinical efficiency. We explore various themes that were generated by pill counting personnel and supervisors in eSwatini. Studies that directly focus on pill counting with a plastic dish and spatula are scarce, but our interviews brought an important issue into focus: staff personnel may frequently be inaccurately counting prescriptions and these errors are not always remedied. The net gain or loss of pills from inaccurately counting prescriptions at a facility level is unclear, but both overcounted and undercounted prescriptions have negative downstream effects. A systematic review examining dispensing errors in the UK, Brazil, USA and France found that prescription error rates between countries varied between 0.015% and 33.5% of prescriptions, and of the nine dispensing error types examined, the wrong quantity was consistently one of the most common types of errors (Aldhwaihi et al., 2016). While pill quantity accuracy is currently unknown in eSwatini, our interviews uncovered instances in which some staff knowingly did not correct mistakes. This suggests that prescription quantity mistakes may be commonplace. Quantitative studies of pill counting accuracy are needed to determine the true extent of this problem. Participants cited feeling pressed for time, being distracted and experiencing frequent interruptions as causes for inaccuracies, and these findings are consistent with research into causes of dispensing errors (Flynn et al., 1996; Desselle, 2005). However, to the best of our knowledge, we have not found in the literature that emotional exhaustion is known as a major cause of pill counting errors. Considering the HIV/AIDS disease burden in eSwatini, the emotional toll that healthcare workers endure is unsurprising. Research in Malawi, a similarly resource-limited country, showed that healthcare worker burnout was especially prevalent in highly endemic HIV environments (Kim et al., 2018). Similarly, in a systematic review examining healthcare worker burnout, nurses in South Africa, with corresponding working environments to eSwatini, had the highest emotional exhaustion score of all countries examined, with a 99.6% prevalence of moderate-to-high emotional exhaustion (Engelbrecht et al., 2008). Not only are patients and intense time constraints at work a source of stress, but in endemic areas, it is reasonable to believe that healthcare workers themselves are impacted by HIV/AIDS and face at-home stressors similar to their patients, such as family member deaths and psychosocial factors related to poverty. Although we did not quantify burnout, our research suggests that mental fatigue was ubiquitous among personnel and that this was identified as a potential contributing factor to pill counting errors. One consequence of pill counting errors is an important effect on the monitoring of patient medication adherence. There are many methods to monitor medication adherence, but according to Liu et al (2001), checking by periodic pill counts is considered at the lower end of sensitivity. Due to its inherent simplicity, periodic checks is the primary method for adherence method in eSwatini. While it is considered an objective measure of adherence, a major drawback is the assumption that the quantity of pills originally dispensed is accurate. Pill counting errors have been identified previously as one of the many reasons for apparent and actual patient non-compliance with medication (Lam and Fresco, 2015). However, the prevalence of such errors has not been thoroughly researched in low-resource settings. Pill counting errors causing apparent, but not actual, non-adherence should be taken seriously in eSwatini facilities for several reasons. First, false accusations of non-adherence can significantly undermine trust between patient and provider and it is well established that patient–provider trust is a critical component to improving patient outcomes (Kramer and Cook, 2004). Second, adherence checks through pill counts are supposed to be an objective measure, but when a patient tells a healthcare worker that they received extra or not enough pills to begin with, this could weaken the healthcare workers perceived usefulness of adherence checking. Another consequence of inaccurately counted pills is challenges related to maintaining medication stocks. Medication stock outs in health facilities in eSwatini are extremely common (Shabangu and Suleman, 2015). Erroneously providing more pills than prescribed exacerbate these stock outs with significant consequences. In a study that looked at non-communicable disease medication stock outs from the patient’s perspective in a hospital in eSwatini over a 6-month period, 73% of patients reported not receiving all of their prescribed drugs at every visit because of medication availability (Shabangu and Suleman, 2015). In addition to increasing the financial burden to the health system, wasted pills and subsequent shortages cause significant financial burden to patients. If patients resort, by necessity, to buying medicines in the private sector instead of the public sector, according to Shabangu and Suleman (2015), the same medications often cost 10–50 times more, an unfeasible amount for most eSwatini people. Furthermore, stock outs have detrimental effects on health outcomes and patients themselves perceive stock outs as poor quality of care, weakening their trust in the health system (Agyepong et al., 2017). A common method of reducing stocks outs and keeping an ample supply of medicines is comparing the expected vs the actual pills leftover in opened bottles. In the case of a mismatch, the facility would have no way of knowing if pills were poorly counted, lost or stolen. Thus, inaccurately counting pills can disrupt drug management processes, because medicines are often shipped to match the predicted patient load of the following weeks. In terms of clinic quality and efficiency, there is a significant shortage of human resources for health in eSwatini, making healthcare workers’ available free time extraordinarily valuable (eSwatini Ministry of Health, 2012). Pill counting with a plastic dish and spatula is tedious and inefficient. And at present, there is no solution for enhancing efficiency in counting prescriptions other than assigning more personnel to the task. There is also an inherent tension between counting quickly and counting accurately. In facilities where everyone shares the burden of pre-packing pills, pill counting has the potential to impact every task within the facility because time counting pills is often time spent away from another activity. This is especially concerning for tasks associated with patient care. Our research shows that a primary concern for nurses and clinical supervisors is that they spent an excessive amount of time counting pills and that this was an unsatisfactory use of their time. For these highly trained nurses, time spent counting pills is time subtracted away from other imperative clinical duties, such as managerial tasks or patient interactions. The WHO has recommended that one approach for health facilities to make more use of available staff personnel is the use of task shifting (Glenton et al., 2013). This method of utilizing untrained staff personnel to undertake various clinical roles is widely used in eSwatini, as seen when cleaners and other lay health workers help count pills, but we observed that cleaners often felt unprepared to dispense medications to patients and were untrained to answer questions. Regardless of who is counting and dispensing the medication, limitations on time shifts the focus of pill counting away from accuracy and towards speed. The problems resulting from staff shortages are further exacerbated by scaling up ART availability. Health facilities do not currently have the capacity to absorb new patients as patients attempt to access care at the health facility level (Van Damme et al., 2008). Ultimately, the excessive amount of effort and time that healthcare workers channel into counting pills has an impact on patient care. This is particularly evident in the extended wait times when staff personnel have run out of pre-counted pills and have to count out prescriptions on demand. Previous studies show that the availability of pre-packed medications compared with counting in the moment reduces waiting time at the dispensary by 50% (Yeboah-Antwi et al., 2001). Long waiting times are correlated with the perception that the quality of care is low, resulting in decreased utilization of care (Yeboah-Antwi et al., 2001). Especially in eSwatini, with its high disease burden, it is reasonable to conclude that poor perceptions of quality of care could translate into increased risks of morbidity and mortality. It is noteworthy that our research suggests a connection between the pill counting burden and the staff’s perception of diminished quality of care. Our research did not quantify the quality of care, nor did we interview patients, but we discovered that HIV counsellors, TB counsellors and nurses perceived their ability to manage patient care was associated with their pill counting workload. This is important because the complexity of HIV and TB medication regiments makes patient-centred education and management crucial for successful outcomes (Munro et al., 2007; Wawrzyniak et al., 2013). Conclusion The WHO’s Sustainable Development Goals highlight the recent global emphasis of quantity ‘and quality’ of health care in low-resource environments. While tremendous growth in the access of ART and other essential medications has occurred in the past decade in eSwatini, our research shows that the onerous task of dispensing these life-saving medications has emerged as a challenge for health facilities. In the context of eSwatini, pill counting is usually a zero-sum activity: time spent counting pills is time spent away from patient care or clinical duties. This study of pill counting personnel and health facility supervisors highlighted many areas in which pill-dispensing activities are suboptimal. Further research is needed on the quantitative accuracy of prescriptions dispensed in eSwatini and on cost-effective and facility-appropriate solutions to alleviate pill counting burdens. Pill counting in eSwatini is a bottleneck to the delivery of health care, which may be detracting from the quality of care that health facilities would otherwise be potentially able to deliver. Pill counting is not an insular task confined to a particular personnel’s workload but instead is carried out by many healthcare workers with widespread and complex implications for the health facility, healthcare workers and patients. Acknowledgements We thank the staff of Management Sciences for Health in eSwatini for their support. We are appreciative of the eSwatini Ministry of Health for facilitating this study. We particularly would like to thank the study participants who made time in their busy clinical schedules to participate in this study. This study was funded by Opportunity Solutions International (http://www.opportunitysolutions.org), a 501(c)3 non-profit organization focused on health innovation and research in resource-limited settings. Conflict of interest statement. Opportunity Solutions International has a patent application pending for a pill counting device for resource-limited settings called SAFEcount. J.D.K. is the inventor of SAFEcount and is co-owner of the intellectual property for SAFEcount. All other authors declared no conflict of interest. Ethical approval. This study received Institutional Review Board approval from Partners Human Research Committee at Massachusetts General Hospital (IRB # 2016P001099/MGH) and eSwatini’s Scientific and Ethics Committee at the Ministry of Health (IRB # 000-9688). References Agyepong IA , Sewankambo N , Binagwaho A et al. . 2017 . The path to longer and healthier lives for all Africans by 2030: the Lancet Commission on the future of health in sub-Saharan Africa . The Lancet 390 : 2803 – 59 . Google Scholar Crossref Search ADS WorldCat Aldhwaihi K , Schifano F , Pezzolesi C , Umaru N. 2016 . A systematic review of the nature of dispensing errors in hospital pharmacies . Integrated Pharmacy Research & Practice 5 : 1 – 10 . Google Scholar Crossref Search ADS PubMed WorldCat Braun V , Clarke V. 2006 . Using thematic analysis in psychology . Qualitative Research in Psychology 3 : 77 – 101 . Google Scholar Crossref Search ADS WorldCat Desselle SP. 2005 . Certified pharmacy technicians’ views on their medication preparation errors and educational needs . American Journal of Health-System Pharmacy 62 : 1992 – 7 . Google Scholar Crossref Search ADS PubMed WorldCat Engelbrecht MC , Bester CL , Van den Berg H , Van Rensburg HCJ. 2008 . A study of predictors and levels of burnout: the case of professional nurses in primary health care facilities in the free state . South African Journal of Economics 76 : S15 – 27 . Google Scholar Crossref Search ADS WorldCat eSwatini Ministry of Health. 2012 . Human Resources for Health Strategic Plan 2012–2017. http://www.nationalplanningcycles.org/sites/default/files/planning_cycle_repository/swaziland/human_resources_for_health_strategic_plan.pdf, accessed 7 February 2020. eSwatini Ministry of Health. 2013 . Swaziland Pharmaceutical Strategic Plan (2012–2016): Baseline Survey. http://apps.who.int/medicinedocs/documents/s21519en/s21519en.pdf, accessed 7 February 2020. Flynn EA , Barker KN , Gibson JT et al. . 1996 . Relationships between ambient sounds and the accuracy of pharmacists’ prescription-filling performance . Human Factors 38 : 614 – 22 . Google Scholar Crossref Search ADS PubMed WorldCat Fung EY , Leung B , Hamilton D , Hope J. 2009 . Do automated dispensing machines improve patient safety? The Canadian Journal of Hospital Pharmacy 62 : 516 – 9 . Google Scholar Crossref Search ADS PubMed WorldCat Glenton C , Colvin CJ , Carlsen B et al. . 2013 . Barriers and facilitators to the implementation of lay health worker programmes to improve access to maternal and child health: qualitative evidence synthesis . Cochrane Database of Systematic Reviews 10 : CD010414 . WorldCat Joint United Nations Programme on HIV/AIDS (UNAIDS). 2017 . UNAIDS DATA 2017. http://www.unaids.org/sites/default/files/media_asset/20170720_Data_book_2017_en.pdf, accessed 7 February 2020. Kieny MP , Bekedam H , Dovlo D et al. . 2017 . Strengthening health systems for universal health coverage and sustainable development . Bulletin of the World Health Organization 95 : 537 – 9 . Google Scholar Crossref Search ADS PubMed WorldCat Kim MH , Mazenga AC , Simon K et al. . 2018 . Burnout and self-reported suboptimal patient care amongst health care workers providing HIV care in Malawi . PLoS One 13 : e0192983 . Google Scholar Crossref Search ADS PubMed WorldCat Kramer RM , Cook KS. 2004 . Trust and Distrust in Organizations: Dilemmas and Approaches . New York, NY: Russell Sage Foundation . Google Preview WorldCat COPAC Lam WY , Fresco P. 2015 . Medication adherence measures: an overview . BioMed Research International , 2015 : 217047 . Google Scholar PubMed WorldCat Liu H , Golin CE , Miller LG et al. . 2001 . A comparison study of multiple measures of adherence to HIV protease inhibitors . Annals of Internal Medicine 134 : 968 – 77 . Google Scholar Crossref Search ADS PubMed WorldCat Management Sciences for Health. 2012 . Ensuring Good Dispensing Practices. http://apps.who.int/medicinedocs/documents/s19607en/s19607en.pdf, accessed 7 February 2020. Munro SA , Lewin SA , Smith HJ et al. . 2007 . Patient adherence to tuberculosis treatment: a systematic review of qualitative research . PLoS Medicine 4 : e238 . Google Scholar Crossref Search ADS PubMed WorldCat Pilchik R. 2000 . Pharmaceutical blister packaging, part I: rationale and materials . Pharmaceutical Technology 24 : 68 . WorldCat Shabangu K , Suleman F. 2015 . Medicines availability at a Swaziland hospital and impact on patients . African Journal of Primary Health Care & Family Medicine 7 : 1 – 6 . Google Scholar Crossref Search ADS WorldCat Ugochukwu CG , Uys LR , Karani AK , Okoronkwo IL , Diop BN. 2013 . Roles of nurses in Sub-Saharan African region . International Journal of Nursing and Midwifery 5 : 117 – 131 . Google Scholar Crossref Search ADS WorldCat United Nations. 2012 . Millennial Development Goal 8. The Partnership for Development: Making Rhetoric a Reality. https://www.who.int/medicines/mdg/mdg8report2012_en.pdf?ua=1, accessed 7 February 2020. Van Damme W , Kober K , Kegels G. 2008 . Scaling-up antiretroviral treatment in Southern African countries with human resource shortage: how will health systems adapt? Social Science & Medicine (1982) 66 : 2108 – 2121 . Google Scholar Crossref Search ADS PubMed WorldCat Wawrzyniak AJ , Ownby RL , McCoy K , Waldrop-Valverde D. 2013 . Health literacy: impact on the health of HIV-infected individuals . Current HIV/AIDS Reports 10 : 295 – 304 . Google Scholar Crossref Search ADS PubMed WorldCat World Bank. 2017a . Population, Total. https://data.worldbank.org/indicator/SP.POP.TOTL?locations=SZ, accessed 7 February 2020. World Bank. 2017b . Prevalence of HIV, Total (% of Population Ages 15–49). https://data.worldbank.org/indicator/SH.DYN.AIDS.ZS?locations=SZ, accessed 7 February 2020. World Bank. 2017c . Rural Population (% of Total Population). https://data.worldbank.org/indicator/SP.RUR.TOTL.ZS?end=2018&locations=SZ&start=1960&v+iew=chart, accessed 7 February 2020. World Bank. 2017d . The World Bank in eSwatini. https://www.worldbank.org/en/country/eswatini/overview, accessed 7 February 2020. World Health Organization. 2016 . Health Workforce Requirements for Universal Health Coverage and the Sustainable Development Goals. https://apps.who.int/iris/bitstream/handle/10665/250330/9789241511407- eng.pdf?sequence=1, accessed 7 February 2020. Yeboah-Antwi K , Gyapong JO , Asare IK et al. . 2001 . Impact of prepackaging antimalarial drugs on cost to patients and compliance with treatment . Bulletin of the World Health Organization 79 : 394 – 9 . Google Scholar PubMed WorldCat Yip W , Hafez R. 2015 . Improving Health System Efficiency. Reforms for Improving the Efficiency of Health Systems: Lessons Learned from 10 Country Cases. https://apps.who.int/iris/bitstream/handle/10665/185989/WHO_HIS_HGF_SR_15.1_eng.pdf;jsessionid=3E28272C22DC3BC11EBC2E345DA70F48?sequence=1, accessed 7 February 2020. © The Author(s) 2020. Published by Oxford University Press in association with The London School of Hygiene and Tropical Medicine. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
TI - The impact of pill counting on resource-limited health facilities: a thematic qualitative analysis in eSwatini
JF - Health Policy and Planning
DO - 10.1093/heapol/czaa007
DA - 2020-05-01
UR - https://www.deepdyve.com/lp/oxford-university-press/the-impact-of-pill-counting-on-resource-limited-health-facilities-a-bo1Aj5oXwh
SP - 1
VL - Advance Article
IS - 
DP - DeepDyve
ER -