Availability and use of therapeutic interchange policies in managing antimicrobial shortages among South African public sector hospitals

Background: Therapeutic interchange policies in hospitals are useful in dealing with antimicrobial shortages and minimising resistance rates. However, the extent of antimicrobial shortages and availability of therapeutic interchange policies is unknown among public sector hospitals in South Africa. Objective This study aimed to ascertain the extent of antimicrobial shortages among public sector hospitals, the presence of current therapeutic interchange policies and the role of pharmacists in the process. Setting Public sector hospitals in South Africa. Methods: A quantitative and descriptive study was conducted with a target population of 403 public sector hospitals. Data were collected from hospital pharmacists using an electronic questionnaire, administered via SurveyMonkeyTM. Main outcome measure Prevalence of public sector hospitals with antimicrobial shortages over the past six months and the prevalence of hospitals with therapeutic interchange policies. Results: The response rate was 33.5%. Most (83.3%) hospitals had experienced shortages in the previous six months. Antimicrobials commonly reported as out of stock included cloxacillin (54.3%), benzathine benzylpenicillin (54.2%), erythromycin (39.6%) and ceftriaxone (38.0%). Reasons for shortages included pharmaceutical companies with supply constraints (85.3%) and an inefficient supply system. Only 42.4% had therapeutic interchange policies, and 88.9% contacted the prescriber when there was a need for substitution. Conclusions: Antimicrobial shortages are prevalent in South African public sector hospitals with penicillins and cephalosporins being the most affected. Therapeutic interchange policies are not available at most hospitals. Effective strategies are required to improve communication between pharmacists and prescribers to ensure safe, appropriate and therapeutically equivalent alternatives are available.

worldwide, enhanced by high prevalence rates of HIV, malaria and tuberculosis [1][2][3], with antimicrobials crucial to reduce the burden of communicable diseases [4]. On one hand, there are concerns with inappropriate use of antibiotics leading to increasing rates of antimicrobial resistance (AMR) limiting therapeutic options, which have resulted in antimicrobial stewardship programmes (ASPs) across countries acknowledging the challenges [5][6][7][8]. However, shortages of antimicrobials are also a challenge in Africa [9][10][11], potentially resulting in the use of antibiotics that place patients at greater risk of Clostridium difficile infections as well as increased AMR [12,13].
The medicine shortage situation in South Africa is also a concern receiving media attention [14][15][16].
As a result, national monitoring programmes have been put into place in South Africa with the support of the National Department of Health (NDoH) surrounding improved supplies management as well as the 'Stop Stockouts' initiative [17,18]. Despite these initiatives, shortages are still being experienced in some provinces in South Africa [16,18,19].Shortages of antimicrobials are regarded as a public health emergency due to the necessity to expedite treatment in the case of an infection and because escalating AMR rates limit therapeutic options for many pathogens [20,21].
Where shortages of antimicrobials exist, therapeutic interchange policies can potentially help ensure that substitution is not haphazard or inappropriate [22]. Pharmacists can play a crucial role in the development and implementation of therapeutic interchange policies in hospitals combined with other key stakeholder groups, building on activities within Pharmacy and Therapeutic Committees (PTCs) as well as ASPs [8,23,24]. Whilst the effective functioning of PTCs is improving in South Africa and ASPs are evolving [23,24], there are concerns regarding the extent of therapeutic interchange policies among public hospitals in South Africa especially regarding antimicrobials. The NDoH published a policy in 2017 to provide guidance for the placement of medicines in classes to support therapeutic interchange programmes where pertinent [25]. However, the policy only provides guidelines on required procedures when switching a patient from one medicine to another within the same therapeutic class, but not on therapeutic interchanges where this is not possible such as antimicrobials. As a result, there is a need to build on this initiative.
Consequently, this study sought to identify and describe current therapeutic interchange policies in the event of drug shortages among public sector hospitals in South Africa following the NDoH initiative. Antimicrobials were chosen in view of the extent of infectious diseases in sub-Saharan Africa, with public sector hospitals chosen as they offer health services to approximately 80% of the population in South Africa [17]. The findings can be used to guide future policies in South Africa.

Materials And Methods
We used a descriptive survey design and a quantitative research approach. All public sector hospitals including district, regional and tertiary/academic hospitals (n=403) in the nine provinces of South Africa were targeted for participation. One pharmacist from each hospital, in most instances the 'Drug Controller', 'Procurement Pharmacist', or 'Pharmacy Manager', were purposively selected to complete the questionnaire. In hospitals where none of these were available, a qualified pharmacist or community service pharmacist was requested to participate.
Data were collected from March to July 2018. The survey was conducted using a 23 item electronic questionnaire administered via SurveyMonkey™. A questionnaire link was sent via email or fax to pharmacists who agreed to take part with anonymous completion. The questionnaire was based on the published literature [11,22,[26][27][28][29] and followed by expert review. It was divided into three sections: demographics, antimicrobial shortages and therapeutic interchange. The questionnaire also included information on current PTC and ASP activities in the hospitals.
Data were exported from SurveyMonkey™ to Microsoft Office Excel™ and cleaned prior to analysis using the Statistical Package for the Social Sciences (SPSS) version 25. Open-ended responses were typed into MS Excel™, and relevant categories were created to allow for counting of responses.
Descriptive statistics were used to summarise data using means with standard deviation (SD) and frequency counts with percentages. Antibiotics were classified by ATC class [30]. All analyses were conducted at a 95% confidence interval.
Ethical clearance for the study was granted by the Sefako Makgatho University Research Ethics Committee (SMUREC/P/269/2017:PG), and permission obtained from the NDoH.

Response rate and demographic details
Of the 403 hospitals, approval to conduct the study was received for 346 hospitals (Table 1). Of these 346 hospitals, 81 facilities could not be reached using email addresses and 11 declined to participate.
A total of 85 responses were finally received, giving a response rate of 33.5% (Table 1). More than half of the respondents were female (63.5%), with a mean age of 38 years (SD: 9.8) ( Table   2). The mean number of years in practice was 9.3 years (SD: 7.42), with 62.2% having less than 10 years experience. Almost half (48.2%) were pharmacy managers, with most working at district hospitals ( Table 2).

Antimicrobial shortages
The majority of respondents (83.3%) indicated they had experienced antimicrobial shortages at their institutions in the previous six months. Most shortages were for antibiotics particularly penicillins, with a limited number in the antiviral (acyclovir) and antifungal (amphoteracin B, clotrimazole) classes (Table 3). Most of the shortages exceeded 40 days. a Total number of respondents who indicated the antimicrobial as either available or not available.

Reasons for antimicrobial shortages
The majority of respondents (85.3%) stated that supply problems with pharmaceutical companies were the main contributor to shortages ( Table 4). The impact included dispensing later generation (84.4%) and more expensive alternatives (64.9%). The procedures for reporting medicine shortages included sending weekly reports to the central/ provincial office or district pharmacist (Table 4).

Role of the pharmacist in therapeutic interchange
More than a third of the respondents (37.1%) said the pharmacist should communicate with prescribers regarding suitable therapeutic options when there are shortages, as well as facilitate the development of therapeutic interchange policies (22.9%) to improve subsequent care.

Discussion
We believe this is the first comprehensive study among public sector hospitals in South Africa to review the current situation regarding therapeutic interchange policies following NDoH guidance [25].
The response rate of 33.5% despite sending up to 12 reminders was similar to one study in the US (40%) [31]; however, higher than studies in Europe and another study in the US with response rates of 22% and 13% respectively [29,32]. Pharmacy managers provided the highest number of responses (Table 2), similar to a study conducted in the US [33].
The majority of participating hospitals had experienced antimicrobial shortages in the preceding six months, similar to studies conducted in Australia, Europe and the US [32,34]. Shortages were either reported to the pharmacy manager, the PTC, hospital management or the provincial office by the pharmacists, similar to the findings in a US survey [31]. It is important that shortages are reported to the relevant authorities and immediately communicated to prescribers so that pertinent strategies can be urgently identified and implemented.
The majority of shortages reported were for penicillins followed by cephalosporins (Table 3), similar to findings in the US, Europe and Australia [10,11,32,34]. Most of the shortages lasted more than 40 days. This is a concern although shortages between one week and over a year have been seen in Europe and the US [11,32]. Pharmaceutical companies with supply or capacity problems, an inefficient supply system, poor stock control and financial resources were the main reasons for shortages (Table 4). However, new initiatives have recently been introduced in South Africa to help improve stock control, and we will be researching their impact in future studies [17].
Most pharmacists reported they resorted to dispensing later generation antimicrobials or more expensive alternatives when faced with shortages, similar to studies in Europe and Australia [32,34].
This though may potentially hinder AMS efforts as shortages of one antimicrobial can result in shortages of others used for substitution, potentially negatively impact on patient outcomes [35].
Encouragingly, the majority of the respondents said their hospitals had active PTCs and AMSCs. This is welcomed especially as one of the aims and objectives of the South African National Drug Policy was to establish and strengthen PTCs in all hospitals [23,24,36]. However, of concern was that less than half of the respondents (42.4%) stated their hospitals had therapeutic interchange policies. Of these, only nine had policies or substitution guidelines from their institutional PTCs. This is consistent though with a recent survey conducted among public sector hospitals in South Africa which found that the development and implementation of guidelines other than the formulary is not a primary function of hospital PTCs [23]. In addition, only 33% of participants referred to the NDoH notice to develop their therapeutic interchange policies. This needs to be addressed as it is difficult to adhere to and revise guidelines that are not documented and actively communicated.
Half of the respondents stated that therapeutic interchange policies were the responsibility of PTCs, with PTCs generally considered the most ideal setting for such programmes which is encouraging as they can ensure that any guidelines are based on scientific evidence and any national Essential Medicine List [17,[37][38][39]. Several actions were reported as part of the pharmacist's practice during substitution (Table 5) including communication with other health professionals and providing information on available alternatives. This is also important as decisions should be made in collaboration with all key stakeholders [35]. Encouragingly, very few respondents stated they substituted without consulting the prescriber, with 92.5% stating that pre-consultation with the prescriber was a requirement before substitution in their hospital. This is important for building trust among teams. The opposite was observed in the US where 67% of institutions performed automatic substitutions preapproved by PTCs [31]. More than a quarter of respondents sent the patient back to the prescriber when a prescribed medicine was unavailable, implying that shortages were not always communicated in a timely manner. This needs to be addressed as it is important that pharmacists actively communicate with prescribers and help jointly develop therapeutic interchange policies to optimize patient care and minimize patient inconvenience [22,40].
Encouragingly, keeping a record of the interchange was undertaken by more than half of the respondents. Encouraging as well was that respondents emphasised the importance of communication with prescribers, active participation in research related to substitution decisions as well as facilitation of therapeutic interchange policy development (Table 6) in going forward. We will be building on this in future studies. Finally, some pharmacists highlighted the importance of effective communication channels between the National/Provincial Department of Health and health facilities, which is consistent with the WHO's recommendation to scale up systems at a national level that collect and monitor data on medicine availability for better evidence-based policy making [41].
The study had several limitations. We acknowledge there was a relatively low response rate, more than half of the respondents had less than 10 years experience, and the responses were selfreported. It is also likely that not all shortages were accounted for due to recall bias. There also appeared to be differences in the interpretation of therapeutic interchange policies among the pharmacists taking part. Despite these limitations, we believe the study was robust providing an insight into the current status of antimicrobial shortages and interchange policies among public sector hospitals in South Africa.

Conclusions And Recommendations
From our findings, it is evident that most South African public sector hospitals experience antimicrobial shortages, particularly for penicillins and cephalosporins. These shortages may last in excess of 40 days, which can affect service delivery, AMS efforts and patient outcomes. Of concern is that the practice of therapeutic interchange is currently not that common in South Africa and not entirely understood by pharmacists working in public sector hospitals. Efforts should be made to train PTC members on how to develop evidence-based policies to manage medicine shortages in their institutions, with pharmacists playing a key role based on their training. There should also be increased communication with depots and ordering early considering depot lead times.
We also believe based on our findings that clear processes for the effective communication of Furthermore, therapeutic interchange policies should be considered for implementation at various levels of care as they take into consideration the spectrum of activity, cost and associated adverse drug reactions of different antimicrobials. This can be part of ASP programmes. We hope our findings are of interest to other low to medium income countries experiencing challenges with medicine shortages especially antimicrobials in hospitals. Availability of data and material

Abbreviations
The datasets during and/or analysed during the current study are available from the corresponding author on reasonable request.

Competing interests
The authors declare that they have no competing interests.
Funding AKC conducted the study as part of a master's degree qualification and received student tuition funding from the National Research Foundation. The study was self-funded, apart from the license for the use of SurveyMonkey™, which was funded through a National Research Foundation grant.
Authors' contributions AKC, MM and JCM conceptualised and designed the study. AKC developed the data collection tool with contribution from MM and JCM. AKC collected and analysed the data. All authors participated in the interpretation of the data. AKC and BG wrote the first draft of the manuscript. All authors commented on the manuscript and agreed on the final version.