Workforce interventions to improve access to emergency contraception pills: a systematic review of current evidence in low- and middle-income countries and recommendations for improving performance

Background Emergency contraceptive pills (ECP) are one of the 13 essential commodities addressed by the UN Commission on Life-Saving Commodities for Women and Children. Although ECP have been available for 20 years, a number of barriers still limit women’s access ECP in low and middle-income countries (LMIC). The workforce who prescribe or dispense ECP are diverse reflecting the varied contexts where ECP are available across the health, commercial and justice sectors and in the community. No reviews currently exist that examine the roles and experiences of the workforce that provide ECP in LMIC. Method We present a narrative synthesis of research to: identify provider factors that facilitate and constraint access to ECP; assess the effectiveness of associated interventions and; explore associated health system issues in LMIC. A search of bibliographic databases, meta-indexes and websites was undertaken to retrieve peer reviewed and grey literature. Literature was screened and identified documents examined to appraise quality. Results Thirty-seven documents were included in the review. Studies focused on formal health workers revealing knowledge gaps concerning the role of private sector and non-health providers who increasingly provide ECP. Data from the findings section in the documents were coded under 4 themes: provider knowledge; provider attitudes and beliefs; provider practice and provider training. The analysis revealed provider knowledge gaps, less than favourable attitudes and practice issues. The findings provide limited insight into products prescribed and/or dispensed, the frequency of provision, and information and advice offered to consumers. Pre and in-service training needs were noted. Conclusion As the provision of ECPs shifts from the clinic-based health sector to increasing provision by the private sector, the limited understanding of provider performance and the practice gaps revealed in this review highlight the need to further examine provider performance to inform the development of appropriate workforce interventions. A standardized approach to assessing performance using agreed outcomes measures may serve to ensure a systematic way forward that is inclusive of the diverse workforce that deliver ECP. Recommendations are outlined to enhance the performance of providers to improve access to ECP. A framework is offered to help guide this process with indicators.


Background
Universal access to sexual and reproductive health and rights is a necessary part of healthy societies. Globally, 222 million women who want to prevent pregnancy are not accessing effective, modern methods of contraception. As a result, each year there are approximately 86 million unplanned pregnancies, 33 million unplanned births [1] and 20 million unsafe abortions [2]. Adolescent women aged 15-19 years give birth to 15 million babies each year with over 90% in low and middle-income countries (LMIC) [3]. Complications from pregnancy and birth are among the leading causes of death for young women [4] and have been linked to poor access to health services, information and care [5]. A competent, motivated and well-managed workforce, as part of a robust health system, is essential to the delivery of evidence-based packages of care for women to improve reproductive and maternal health outcomes [6,7].
Emergency contraceptive pills (ECP) provide women with a safe and fairly effective opportunity to prevent pregnancy after unprotected intercourse. ECP is one of the 13 essential commodities that are addressed by the UN Commission on Life-Saving Commodities for Women and Children (UNCoLSC) [8]. Despite being available for nearly 20 years, a number of barriers still limit women's access to ECP [9]. The workforce who prescribes or dispense ECP are diverse and reflect the varied contexts where ECP is available across the health, commercial and justice sectors, as well as in the community. In both the public and private health sector at the primary, secondary and tertiary levels, doctors, pharmacists, nurses, midwives and paramedical staff such as nursing assistants can provide access to ECP. At the community level, formal and lay health workers distribute contraceptives. In Bangladesh for example, family welfare assistants supply contraceptives during home visits and volunteer female community health workers (CHW), who reside in rural villages and urban neighbourhoods distribute condoms and contraceptive pills to women [10]. As part of a global trend to move ECP from a prescription to an over the counter (OTC) product, the commercial private sector is playing an increasingly important role. ECP are widely available in pharmacies [11] and sold by pharmacists and retail staff also known as drug vendors or patent medicine vendors [12]. Other providers in the commercial sector include drug sellers such as patent medicine vendors or retailers in Nigeria who have no formal pharmacy training and sell pharmaceutical products for profit. These providers are usually the primary source of drugs, particularly for the poor [12]. Other providers have been trained to distribute ECP including the police in situations where women have been sexually assaulted [13].
Insight into the availability of these providers, their practice and the challenges they face may offer an understanding of the factors that influence access to and demand for ECP. However the diversity of these providers does not lend itself to current measures of access to workforce and therefore contraceptive coverage because measures such as the Health Workers' Reach Index (HWRI) or health worker density ratio focus on doctors, nurses, and midwives. These measures ignore the roles other providers play in providing access to ECP and the complexity of the situation can be illustrated by a capacity project study where African countries that have the same HWRI scores can have different contraceptive prevalence rates [14]. No systematic review currently exists that examines the roles and experiences of this diverse workforce that provide ECP in LMIC.
In order to provide evidence to contribute to strengthening ECP delivery channels and developing policy guidance and tools to expand access to ECP particularly among vulnerable groups, we undertook a narrative synthesis of current research focusing on providers of ECP. The aim of this review was to identify workforce opportunities and interventions in LMIC that can facilitate increased access to ECP. The results of this review and related reviews into consumer and service delivery experiences will provide the necessary evidence to assist the ECP Technical Reference Team to help carry forward UNCoLSC recommendations at the global and national levels [15].

Methods
The bibliographic databases (CINAHL, MEDLINE, PubMed, SCOPUS, ProQuest Health & Medical Complete, Web of Science, African Journals On Line), meta-Indexes (Popline, Eldis knowledge services, Reproductive Health Library) and websites of relevant organizations (The Guttmacher Institute, The International Consortium on Emergency Contraception, Population Council) were searched. The mixed nature of the methodologies of studies identified meant that results could not be combined. As a result a narrative synthesis methodology [16] was applied to analyse research papers and reports whose quality has been appraised using acknowledged tools.
A Population, Interventions, Comparators, Outcomes, Study design (PICOS) question was formed to guide this review as per guidelines [17]. The review question was: What provider factors facilitate and constrain access to ECP in LMIC? Outcomes of interest include: provider knowledge; training outcomes, and practices. The review aimed to source descriptive as well as intervention studies that shed light on access to ECP. Observational studies, quasi experimental and non-experimental descriptive studies were considered suitable for inclusion and a systematic search of the primary research literature published from 2003 to 2013 in English and in LMIC was undertaken. Electronic databases and the internet were searched using the Medline Medical subject headings (MeSH) 'Postcoital Contraception' and'Health Services Accessibility' and 'Healthcare Delivery' and 'Contraceptive Distribution' and 'Health Manpower' or 'Health Personnel' and 'Developing Countries' and supplemented by the terms key words 'emergency contraception', or 'emergency contraceptive pills' and 'provider' or 'workforce' or 'nurse' or 'doctor' or pharmacist' using the MeSH terms, abstract and keyword options. For example, a search of ProQuest Health & Medical Complete using the terms ft(emergency contraception) AND ft((provider OR nurse)) AND ft((doctor OR pharmacist)) AND mesh (Postcoital Contraception) peer reviewed documents in English 2003-2013 returned 47 documents. The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines were used to report the review process [18] (see Figure 1). Duplicate records were removed and the remaining records then screened using the PICOS question. Papers without data, those older than 10 years or whose focus was outside of the aim were removed. In the example of the ProQuest Health & Medical Complete database search mentioned above two of the 47 papers were included in the review as their focus was on providers in LMIC and within the date limitations. A similar process was undertaken with the other searches and relevant papers compiled in an Endnote bibliographic database.

Appraisal
Thirty-nine papers were identified as relevant. These papers were appraised by all authors for their quality using the Critical Appraisal Skills Programme (CASP) tool for qualitative research [19] and Pluye et al.'s [20] scoring system was used to assess the non-experimental studies and mixed methods. Research reports were classified as grey literature and were appraised using a checklist developed for this purpose [21]. Two papers were excluded due to the lack of data on ECP [22,23]. Ethical approval was not required for this study as all data is in the public domain.

Data extraction and synthesis
A narrative synthesis approach was undertaken according to Popay [16] allowing qualitative and quantitative data to be examined. The results sections of each of the 37 papers were analysed to identify evidence where providers had helped to increase to ECP in LMIC. A thematic analysis was conducted by the first author using tables and discussed with other authors in order to reach consensus. The relationships within and between studies were explored and coded using QSR NVivo 10 software. The analysis was guided by the World Health Organization (WHO) Health System [24] Building Blocks, a framework that describes the six components of health systems, namely service delivery, health workforce, health information, medical technologies, health financing, and leadership and governance. Concept maps were used to plot patterns and relationships across the themes and subthemes and robustness assessed through critical reflection and discussion among the authors.

Results
Of the 37 papers that were included in review providers included groups of clinically-based health professionals, such as doctors or nurses in hospitals, and health professionals and service staff in retail contexts such as pharmacists and sales people or drug vendors in private pharmacies or shops (see Table 1 for an outline of all provider types in the studies included in this review). Nine studies focused on those working in retail settings, i.e. pharmacist and/or drug vendors, 26 focused on clinic or hospital-based personnel (doctors, nurses) and only two included police.
Geographically, the papers were concentrated in one region: twenty one of the papers were based on studies carried out in Africa (including seven from Nigeria and five from Ethiopia). Fourteen studies were carried out in Asia, with eight of these conducted in India. Two studies in the review were undertaken in Central America.
Descriptive survey designs underpinned the majority of the studies (n = 30) (See Table 2 for a summary of all papers in the review). Data from the findings section in the documents were coded under four themes: provider knowledge; provider attitudes and beliefs; provider practice and provider training.

Provider knowledge
Many studies examined the knowledge of providers including: timing and administration of ECP; its mechanism of action; indications; eligibility; side effects; safety; effectiveness; sites at which it is available; cost; sources of information and factors associated with knowledge. In many of the papers, providers correctly identified at least one emergency contraception product available in their country [25][26][27][28][29][30][31][32][33][34][35]. However, in some countries, some respondents incorrectly identified other products as ECP. Medicine indicated for menstrual irregularities, and the malaria therapy (quinine) were erroneously noted as ECP by medical doctors [33] and drug vendors [36] in Nigeria.

Provider attitudes and beliefs
A lack of understanding of the biological mechanism of action underpinning ECP was noted among many study respondents; some were not aware of the disruption of ovulation or said that ECP blocks the implantation of the fertilized ovum, or were not sure of the mechanism [30,[42][43][44]. Some providers described ECP as an abortifacient [29,37,39,40,42,43,[45][46][47]. In some studies in the review providers incorrectly believed that ECP use could result in congenital abnormalities if a woman was pregnant [29,35,37,39,40,43]. Infertility was a concern of some Nicaraguan study participants [37], while some Ethiopian pharmacists and drug vendors said that ECP could cause uterine and breast cancer, skin pigmentation, abdominal pain, and high blood pressure [35,39].
In some studies, providers expressed concern that ECP use might displace regular contraception use [25,30,35,37,39,47]. Others said they were confident that by counselling their clients they could prevent them from using ECP as a regular method. Health providers in some studies also expressed concern that increasing the availability of ECP would promote earlier sexual debut [35] and lead to unsafe sex [32,48], an increase in STIs and HIV [37,49].
Barriers to prescription and/or dispensing were noted in a few studies, including low consumer demand [26,36], and ethical, legal and religious concerns of the providers [34,36,43]. ECP were regarded by some as delivering a cost saving to the health system through reducing the abortion rate [41] and unwanted child bearing [32], as well as an opportunity to promote regular contraception [47].
Providers in one study were found to be more judgemental of clients who did not use family planning methods other than ECP, nevertheless, most providers said they still served such clients [47]. Some study participants felt that ECP should not be available to unmarried women and adolescents because access would encourage sexual activity [32,37].

Provider practice
Data from the findings of 25 of the papers made reference to the practice of various providers in relation to their prescription or dispensing practices, or interaction with consumers. Four sub-themes emerged describing factors related to provider practice. These sub-themes are: volume of ECP sales, provider characteristics, client characteristics, information and advice. Only two of these studies provided observational data of provider practice from the perspective of "mystery clients" [50,51] while the reminder of the documents relied on provider recall and self-report. In terms of volume of ECP sales, about a quarter of respondents in a Ghanaian study said they provided ECP to clients every day and another quarter provided it several times through the month [47]. In a Nigerian study when providers were asked the number of clients they supplied ECP with each month, the answer ranged from 0-45 women, more than 50% of the respondents served fewer than 10 clients per month [52]. The average estimate per pharmacy staff member in a study in Addis Ababa was 22.5 ECP clients per month [39]. This average was similar to drug vendors surveyed in Nigeria where about 43.2 per cent reported a clientele of fewer    To assess provider knowledge and practice 9/10 providers had heard of emergency contraception, 50% knew the correct time frame for effective use 75% knew that the pills prevent pregnancy; >30% believed that they act as an abortifacient. <30% who had heard of the pills knew they are legal in Nigeria. Of those who had heard about emergency contraception, 58% provided clients with emergency contraceptive pills, 10% of these providers could correctly identify the drug, dose and timing of the first pill in the regimen. < than 1/10 who knew of emergency contraception said they always provided information to clients, whereas ¼ said they never did so     To evaluate the provision of reproductive health information and services to users of ECs by private pharmacists The differences between the control and intervention pharmacies with respect to the provision of additional information on EC and regular family planning services are in the expected direction but statistically insignificant. In contrast, the likelihood of providing information or referral for counselling or testing for sexually transmitted infections or HIV was lower in the intervention than in the control pharmacies but the difference was also not statistically significant.     than 20 people monthly however, one quarter (25.3 per cent) had 20-100 clients, and six per cent had a 100 or more clients monthly [36]. One study noted that the number of packs sold in a month per pharmacy varied from two to 500 packs/month, with a mean of 62 packs every month [53]. ECP is offered over the counter in many countries and women are able to decide for themselves whether they need ECP. The safety profile of ECP, as well as the simplicity of the regimen means that many or most women do not require information when they purchase ECP, other than that provided with the package. However, some women may need or wish more information from providers. The likelihood of providing information and advice to clients purchasing ECP was reported in some studies. In one Nigerian study fewer than one in 10 providers said they always provided information to clients, while one-fourth never provided information to clients obtaining ECP [46]. Nine out of 10 Indian pharmacists surveyed reported simply providing ECP to customers [42]. Two studies involving the use of mystery clients who presented at various pharmacies found that those whose role involved playing an inexperienced client were significantly more likely to be offered additional information on ECP rather than when they presented as experienced users of ECP [50,51], suggesting that providers have the capacity to correctly judge when it is appropriate to offer women purchasing ECP the opportunity to receive more information.
Providers in a number of studies noted a range of ECP consumer characteristics [36,41,42]. The providers interviewed in Khan et al. believed that unmarried younger women or students are more likely to go to pharmacies, whereas married older women are more likely to go to other providers; younger, unmarried women are least likely to go to qualified doctors [42].
Various provider characteristics were examined in some research to identify the factors associated with the provision of ECP including provider knowledge, gender, and place of work, religion, experience and profession. Busy retail outlets were said to limit the provision of information and counselling [50] and stock outs affected provider's ability to dispense [36,42,47,51,52].
Police procedure for processing women who have been sexually assaulted was described by Ethiopian police interviewees in the Kassa et al. study [54]. Women were said to present first at a police station after being assaulted because they require police certification as evidence of assault before being able to receive treatment. A female police officer responsible for dealing with such crimes would interview the victim, file a report and then refer the women to a health facility, usually one that would provide a medical report in a short time frame to aid the investigation [54].

Provider education and training
Providers identified ECP training as a need in four studies [25,32,52,54]. Some surveyed Nigerian and Nepalese providers reporting having received ECP training [41,45] while one Senegalese study found that a considerable proportion of providers (38%) had not received any ECP training [49]. Four studies described training initiatives that had led to improvements in provider knowledge and dispensing practice and more positive attitudes towards ECP [28,47,55,56]. A study of ECP from Guatemala included training for police officers, social workers and health professionals. The findings indicated that these professionals not only shared their ECP knowledge and practice experiences with colleagues in their own organizations but across these sectors [55].

Discussion
This narrative synthesis of 37 documents provided insight into the knowledge, attitudes, practices and training of those who provide ECP to women and men in LMIC. The analysis of the findings of studies retrieved in this review revealed a focus on formal health workers and less on drug vendors, pharmacy staff, lay health workers and providers in the education, justice and social service sectors. This reveals gaps in knowledge about the role of private sector providers, which is particularly acute given the prominent role that the private commercial sector plays in making ECP widely available in LMICs and elsewhere in the world. In general, the literature on the pharmacy sector is scanty, posing a significant gap given that ECP are mostly provided in this sector in the majority of LMIC [57]. None of the papers reviewed focused on providers in humanitarian contexts [58]. Gaps are also visible in the role of health staff in setting such as youth clinics or in schools where adolescents can be provided access to ECP. There is also a dearth of literature reflecting the perspectives of providers in LMIC in the Western Pacific, South America and Central Asia.

Improving human resources for health performance to deliver ECP
Evidence of workforce interventions to improve practice is scarce in the documents reviewed, particularly in terms of the types of workforce support and performance management initiatives that may be required across both the public and private sectors to deliver ECP. Given the number of studies identifying provider knowledge gaps, less than favourable attitudes and issues with practice, investigations into approaches to optimize provider performance to deliver ECP may be useful. This could involve the development of workforce initiatives for the provision of ECP that could be assessed against outcome measures in key human resources for health (HRH) areas such as: policy, regulation and legislation; management systems; education, training and competencies and; consumer engagement. These areas have been described as useful fields for guiding the improvement of workforce performance in reproductive health [59] and are discussed below in relation to indictors for increasing access to ECP. The development and implementation of performance indicators alongside training and structured supervision may not be feasible in context of small, informal or sole operator businesses selling ECP. Other mechanisms for supporting providers in the retail setting, such as pharmaceutical "detailing" or shortcourse format training, may be more feasible and could possibly play an important role in strengthening access to ECP. However, these alternative strategies, while already in use by social marketing organizations and commercial pharmaceutical distributors, have not been evaluated and do not appear in the published literature reviewed here. More formal performance management interventions may still play a role as they could potentially be usefully applied to pre-service training, and to larger chain retail pharmacies and franchised outlets [60] that have experienced rapid growth in LMIC such as India [61].

Workforce and ECP Policy, regulation and legislation
We were not able to identify research concerned with national and district HRH policies that address health workers in the public, private and non-state sectors, job classification systems, registration, certification, or licensing requirements or service functions [62]. While the movement of ECP from a prescription to an over-thecounter (OTC) drug has been a crucial step in expanding access, it appears from our review that nothing exists in the formal literature examining or documenting the impact of this policy change. Future research may be useful to investigate how workforce policy is, or can be best aligned with national reproductive health policies, plans and laws to facilitate access to ECP and how the changing patterns of provision of ECP, from a prescription product in clinic settings to a OTC drug in commercial settings has affected training and information needs for these new cadres of providers. The development of national policy targets in monitoring and evaluation of HRH plans for example may be helpful for measuring progress including in terms of ECP provision. Indicators in this area could include the proportion of workers trained in ECP and if appropriate, the numbers of providers licensed to provide ECP.

Management systems
HRH management refers to staff supply (staff numbers, skills mix and workforce planning), performance management (supervision and motivation), and personnel administration and employee relations (pay and incentives) [63]. There appears to be a lack of research examining the distribution of personnel who dispense ECP in both the public and private health sectors and beyond, and a major gap in the literature regarding the transition of ECP provision from public to private/commercial sector and the impact . This data would provide valuable information for future planning workforce and co-ordination to ensure efficient use of current staff.
Although nurses and midwives were included in the generic studies of health providers in the review only one study focused on nurses. The role of nurses and midwives in reproductive health provision is central and as they comprise the majority of health workers in most countries and it has been argued that the role of midlevel cadres deserves particular emphasis in LMIC where there are chronic shortages of staff [7,64]. Task shifting and task sharing practices involving the delegation of reproductive tasks from doctors to nurses and midwives has been found elsewhere to show promise [60,65,66]. In addition shifting tasks relating to the distribution of ECP to paramedical and community health workers may be a useful strategy as has been found in other research [67]. However these cadres will require training and supportive supervision to carry out these new duties. The current literature does not explore how midwives and nurses can support ECP knowledge (for instance, during contraceptive counselling, post-partum education, etc.) even if they are not typically dispensers of ECPs after the transition to a commercial sector, pharmacy environment.
Research concerning the actual performance of ECP providers other than self-report was lacking in the review, as were standard measures to assess outputs. Calls have been made for measurements of quality in the sexual and reproductive health workforce, in order to provide incentives to deliver quality care and pay for performance [68]. Performance indicators could be developed to include ECP and applied at both health service and provider levels and in the commercial sector. The assessment of provider performance indicators would need to involve quality audit and supervisory processes both of which have been shown to contribute to improve health worker performance and health outcomes [69]. Supervision is also an important ancillary aspect of paying health workers financial incentives for performance that is in line with public health goals [70]. However lacking in the review is research concerning the best practices in providing financial and non-financial incentives to workers to ensure the availability of ECP in the public and private sectors. Vouchers for vulnerable groups may improve access to contraception including ECP [71] but the evidence for pay for performance incentives in LMICs is not strong [72]. It may be that in the commercial sector, sales of ECP at a profitable price can generate adequate motivation for providers to offer ECPs free of barriers however, we did not discover any literature that explores the issue of profit in motivating workforce.
Quality staff performance requires bundles of intervention [73]. This includes audit and supportive supervision in combination with a range of other initiatives such as incentives, job aides and training [69]. Research in high income countries may be transferable to LMIC contexts to improve provider performance to dispense ECP. In Australia, a job aide in the form of a written checklist was found to improve the quantity and consistency of patient assessment but not the quality of advice [74]. Some countries have developed practice guidelines [75] that may provide a direction for others wishing to better support clinicians.

Education and training
In this review no studies were found that examined the pre and in-service curriculum of health providers as it pertained to ECP. This concurs with other studies that have found few evaluations of training initiatives for health and human services providers [76]. Knowledge of useful education and training interventions would help to identify gaps and inconstancies, align curricula with national policy and integrate interprofessional learning opportunities where appropriate. The need for professional education and training has been noted particularly in the commercial sector where turnover of retail staff is high [11]. This need is being partially addressed by the WHO, the United Nations Population Fund (UNFPA) and the United States Agency for International Aid (USAID) Training Resource Package for Family Planning. This package offers essential resources including separate modules for ECP for family planning and reproductive health trainers, supervisors, and program managers (http://www.fp training.org./). Insight into education efforts to address judgemental attitudes towards particular consumers and ECP related stigma could be gained from mental health and pharmacy education where contact-based education was found to reduce mental illness-related stigma [77].
Few studies captured data on the information and advice given to consumers. While in some contexts, information provision may not be necessary as women may have prior knowledge or prefer to read the packet, asking consumers if they wish to have further information would provide an opportunity to offer this if required. While encouraging results were found in two studies [50,51] suggesting that providers could ascertain when women required information, training to hone these skills and best communicate the information women want may be useful.
Four studies in the review included exploration of the knowledge, attitudes and practices of non-health personnel and as previously noted there is little emphasis on private sector providers. Lessons from interventions focused at retailers in other areas such as malaria may be useful. A review of medicine retailers selling malaria medications in sub-Saharan Africa found that on-going supervision and on-the-job-training can contribute to improving the appropriateness of drugs and information provided by these providers [78]. One study in the current review found that training providers in the health, community and justice sectors has positive outcomes in knowledge gains, attitudes and self-reported practice [55]. More knowledge is needed to better understand best practices in improving and training the ECP-capacity of different cadres across sectors. Professional collaboration between health workers and the police has been found to improve access to ECP, however this was the result of a number of initiatives beyond training, including changing policies to enable police officers themselves to provide ECP and introducing reliable supplies, supportive supervision, and revised protocols [13]. Collaboration across the education and community sectors may also be possible through such supportive workplace environments. There is evidence in primary health care contexts that collaboration across sectors can contribute to positive intermediate changes in health knowledge, attitudes, behaviour, and in the environment through new policies, practices and services [79].

Community and consumer engagement
The primary health care approach [80] acknowledges the resourcefulness of communities and the importance of community participation in their own health care. Provider engagement with the community to build relationships, understand and best respond to needs is therefore necessary to increase access to ECP. Provider performance in this area could be gauged through measures of client satisfaction, consumer/patient contacts, and the presence of a formal relationship with community organisations and leaders.
Collaborative community based social marketing initiatives [81] and education could engage community members, consumers and providers in positive ways to promote ECP. Participatory women's groups might improve prove effective in disseminating information about ECP and preventing unwanted pregnancy as they have in the improvement of maternal and neonatal survival [82]. Partnerships with community members to distribute ECP may be beneficial alongside strengthened health system logistics to ensure equitable access. Lessons from other drug distribution efforts may prove useful such as the effective delivery of Intermittent Preventive Treatment of malaria during pregnancy (IPTp) through traditional birth attendants, drug shop vendors, community reproductive health workers and adolescent peer mobilisers in Uganda [83]. However community members who distribute health commodities are also driven by their own personal agendas and interests that can be at odds with public health goals. For example community distributors and drug vendors may have preconceived views about those that seek ECP and how often they wish to procure it, or the stigmatization of certain groups such as unmarried women and adolescents that may affect women's access to ECP. Knowledge of distributors' motivations can help to provide more realistic expectations from such programmes and improve relationships between community distributors, consumers and formal providers [84].

A model for monitoring performance
If provider performance is to be improved in order to increase access to ECP then decisions may need to be made concerning the indicators that should be used to assess such performance and how they need to be tailored to the diverse provider contexts. Workforce performance may need to be considered in relation to other aspects of health systems strengthening included in the WHO health systems framework for action (WHO 2007) and alongside health targets such as Millennium Development Goal (MDG) 5b concerning universal access to reproductive health. The perspective therefore needs to be multidimensional and involve the development of input, process and output level indicators [85] to link planning, implementation and evaluation aspects of ECP provision. This will ideally include the collection of quantitative data concerning the number of ECP actually sold or dispensed, as well as contextual information concerning stocks and supply. However, sales data from the commercial sector are generally not available and therefore it is difficult to ascertain how much ECP is being distributed in a certain country. Some market research data is available for a fee for countries including Cameroon and Kenya [86]. The best proxy for the distribution of ECP may be reports of ECP knowledge and use in population-based surveys such as the Demographic and Health Survey [87] or the Performance Monitoring and Accountability 2020 [88] surveys. Table 3 provides some examples of indicators at input, process and output levels across the six building blocks of the WHO health systems framework. Plotting indicators in a matrix format can provide an opportunity to see how indicators in each of the areas might relate so that appropriate measures can be selected for assessing provider performance.

Study limitations
This review may have been limited by an incomplete identification of research studies; however, efforts were made to search bibliographic databases, meta-indexes and the websites of international agencies working in the field. Additional documents were retrieved as a result however some may have been at a lower quality as not all studies were published in peer reviewed journals. The application of a narrative synthesis methodology to the results of the studies included in this review may have led to a reduction in detail particularly in terms of contextual factors that are relevant to the outcomes of interventions. However we made efforts to maintain detail in this review through rich textual descriptions of the study's findings that provided a narrative across all studies. The graphical elements included in the analysis were useful for identifying patterns and the tabulation of findings enabled structured comparisons where pertinent.

Conclusion
Adequate numbers of well-motivated, managed and competent providers are a critical part of delivering lifesaving commodities to reduce maternal death and provide access to reproductive health to achieve MDG 5 targets. As ECP provision increasingly shifts from the health care clinic sector to the pharmacy sector through changing policies regarding OTC status, there is an urgent need to understand more about who is now providing ECP in LMIC and what kind of training they have and what training would be required to improve performance. Ideally, it would be important to establish and apply evidence-based workforce strategies in an integrated manner to attain optimum performance to improve access to ECP, for a wide range of cadres of staff, including pharmacists and drug vendors. This review reveals considerable knowledge gaps concerning the performance of those who currently prescribe and/or dispense ECP. A focus on developing ways to best support this emerging workforce may improve practice and hence access to ECP for women.