Availability, prices and affordability of essential medicines for treatment of diabetes and hypertension in private pharmacies in Zambia.

OBJECTIVES: To explore availability, prices and affordability of essential medicines for diabetes and hypertension treatment in private pharmacies in three provinces of Zambia.
METHODS: A cross-sectional survey was conducted in 99 pharmacies across three Zambian provinces. Methods were based on a standardized methodology by the World Health Organization and Health Action International. Availability was analysed as mean availability per pharmacy and individual medicine. Median prices were compared to international reference prices and differences in price between medicine forms (original brand or generic product) were computed. Affordability was assessed as number of days' salaries required to purchase a standard treatment course using the absolute poverty line and mean per capita provincial household income as standard. An analysis identifying medicines considered both available and affordable was conducted.
RESULTS: Two antidiabetics and nine antihypertensives had high-level availability (≥80%) in all provinces; availability levels for the remaining surveyed antidiabetics and antihypertensives were largely found below 50%. Availability further varied markedly across medicines and medicine forms. Prices for most medicines were higher than international reference prices and great price variations were found between pharmacies, medicines and medicine forms. Compared to original brand products, purchase of generics was associated with price savings for patients between 21.54% and 96.47%. No medicine was affordable against the absolute poverty line and only between four and eleven using mean per capita provincial incomes. Seven generics in Copperbelt/Lusaka and two in Central province were highly available and affordable.
CONCLUSIONS: The study showed that the majority of surveyed antidiabetic and antihypertensive medicines was inadequately available (<80%). In addition, most prices were higher than their international reference prices and that treatment with these medicines was largely unaffordable against the set affordability thresholds. Underlying reasons for the findings should be explored as a basis for targeted policy initiatives.

Introduction

In line with its pledge to achieve Universal Health Coverage by 2030, the government of the Republic of Zambia is planning to establish a National Health Insurance (NHI) scheme [1]. The corresponding law was passed in 2018 and mandates the National Health Insurance Management Authority to develop a comprehensive health benefits package (HBP), taking special account of the national burden of disease and national public health priorities [1,2].

One such priority area is the growing burden of noncommunicable diseases (NCDs), primarily diabetes mellitus (diabetes) and hypertension [3,4]. Diabetes is a metabolic condition characterized by excessive blood glucose levels [3]. Hypertension is an arterial disease with persistently elevated blood pressure [4]. Both are chronic, progressive diseases and can lead to severe complications including cardiovascular disease and kidney failure and premature deaths if not appropriately managed and controlled [3,4]. Primary prevention, early diagnosis and adequate treatment are key to avoid such dire consequences and associated economic losses [3,4]. However, diabetes and hypertension often go undetected in lower and middle-income countries (LMICs)–including Zambia [3-5], due to prevailing financial and human resource constraints, insidious onset of the diseases and lack of awareness by those affected. With prevalence estimates of 7.6% for diabetes and 41.6% for hypertension, both accounted for almost half of Zambia’s NCDs burden already in 2013 and were additionally among the 10 leading causes of admission to and mortality in public health facilities from 2009 to 2015 [1,5,6]. Data from the central health management information system (HMIS) indicate that the prevalence has been rising considerably since, a trend which is expected to continue in coming decades [7].

Long-term, often lifelong treatment requiring continuous medication is common for diabetic and hypertensive patients, which makes access to essential medicines (EMs) vital for treatment success and, ultimately, survival [3,4]. Effective antidiabetic and antihypertensive medicines with good margins of safety have been developed and been available for many years, such as oral hypoglycaemics and insulin for diabetes [3] and diuretics, beta blockers and angiotensin converting enzyme inhibitors for hypertension [4].

However, EMs for diabetes and hypertension have been found to be intermittent in supply, unavailable and unaffordable in studies across LMICs [8-10]. Zambia has a policy of providing medicines free of charge for public primary health care [11], yet due to resource limitations, medicines are available irregularly. For instance, a 2017 study in 15 public facilities in Lusaka found availability levels of only 44.7% to 58.2% and stock out rates between 69% and 92% for a number of antidiabetics and antihypertensives [12]. When public facilities are unable to dispense required medicines, patients have to purchase them out of their “own pocket” at prevailing prices in the private sector, where availability levels are frequently higher [9,13,14]. Figures for private sector patient prices in Zambia vary widely and no reliable estimates are available, yet they have generally been observed to be high [14,15]. In this light, the 2014 Zambian Household Expenditure and Utilization Survey found that 42% of out-of-pocket payments (OOPs) incurring in rural and urban populations of all income groups were related to medicine expenses [16]. Although these estimates do not relate to diabetes or hypertension in particular, they are nonetheless indicative of the economic burden these chronic conditions can impose on individuals and households [3,4].

With the planned introduction of the NHI, cooperation and complementarity of health service delivery, including medicine provision, between the public and private sector are foreseen [17]. A solid understanding of availability, prices and affordability of antidiabetic and antihypertensive medicines in both sectors is hence essential to enable planning and realizing an effective intersectoral cooperation. There is, however, a paucity of reliable data and very little research has been carried out into the specific scope of the challenges regarding these medicines. Against this background, the aim of this study was to systematically assess availability, prices and affordability of EMs for diabetes and hypertension treatment in private pharmacies in three provinces of Zambia.

Methods

A descriptive cross-sectional survey of availability, prices and affordability of essential antidiabetic and antihypertensive medicines was conducted. Study design and methods were adapted from a standardized methodology for systematic facility-based surveys developed by the World Health Organization (WHO) and Health Action International (HAI) [18].

Survey scope

The three adjacent provinces Central, Copperbelt and Lusaka were purposively selected as survey scope. They account for about 22% of the total surface area and 43% of the total population (2015) in Zambia [19] and are home to 89% (191 of 214) of all private retail pharmacy businesses (hereinafter referred to as “pharmacies”) that have official operating licenses from the Zambian Medicines Regulatory Authority (ZAMRA) [20]. Copperbelt and Lusaka are Zambia’s most urban and densely populated provinces, while Central is more sparsely inhabited and considered predominantly rural [21]. The rationale behind choosing this study area was based on three main components. Firstly, Zambia is geographically vast and has one of the lowest population densities in Sub-Saharan Africa [21], in which cases the applied WHO/HAI methodology recommends conducting a provincial survey or a series of provincial-based surveys [18]. Secondly, 2018 routine data collected from the central HMIS showed that Central, Copperbelt and Lusaka province had the highest utilization of public hospitals by diabetes and hypertension patients seeking treatment of all provinces. Thirdly, they were found to have the highest shares of total household OOPs of all provinces in the 2014 Zambian Household Expenditure and Utilization Survey, a large proportion of which relates to medicine spending [16].

Selection of pharmacies

A sample of 99 pharmacies was surveyed based on the 2018 register of licensed private retail pharmacies published by ZAMRA [20]. Due to the small total number of registered pharmacies in Central province (n = 6), these were surveyed exhaustively. In Lusaka and Copperbelt province, pharmacies were selected by simple random sampling based on methods developed by Measure Evaluation [22]. The applied formula determined the sample size based on the minimum acceptable margin of error of the estimate, the level of confidence desired and the approximate prevalence of the parameter of interest in the facility population [22]. As the total population of pharmacies was predetermined and relatively small in both provinces (N = 149 in Lusaka and N = 36 in Copperbelt province), the resulting n was greater than 5% of the total, and the sample sizes were correspondingly adjusted with a Finite Population Correction factor [22]. Resulting sample sizes were n = 65 for Lusaka and n = 28 for Copperbelt province. Whilst retaining the numbering of the ZAMRA register, pharmacies located in Copperbelt and Lusaka province were extracted and entered into Microsoft Excel spreadsheets. A random number sequence was then created to randomly select pharmacies corresponding to the sample sizes. In addition, 10% of the sample sizes were selected as backup pharmacies; no backups could be selected in Central province.

Selection of medicines

A standardized data collection form was developed to ensure consistency of data collection across pharmacies and provide for maximum data accuracy and reliability. A total of 32 medicines including 7 antidiabetics and 25 antihypertensives were included in the survey. Medicine selection was based on the 2017 WHO Model List of EMs [23] and the 2017 Zambia Essential Medicines List (ZEML) [24] and further cross-checked against the Zambia National Formulary and global diabetes and hypertension guidelines to ensure completeness [25-27]. Selection was finalised in consultation with a practicing Zambian pharmacist and officials from Medical Stores Limited [28].To be included in the final survey list, applied inclusion criteria were that medicines needed to have international reference prices (IRPs) in the Drug Price Indicator Guide by Management Sciences for Health (MSH) [29] and further to be authorised for sale in Zambia based on the 2018 ZAMRA register of market authorisations of medicines for human use [30]. The MSH IRP set is based upon recent procurement prices for generics offered to LMICs by international suppliers and provides a benchmark of reasonable medicine prices [18,29]. Dosage form and strength were specified as per the essential medicine lists (EMLs). Both single-ingredient and fixed-dose combination products were considered, provided that all the requirements for inclusion in the survey were met. Furthermore, as no standard for package sizes has been adopted in Zambia, these were assumed to correspond to a standard treatment course (STC) for chronic diseases, defined as 30 days [18]. To verify effectiveness of the data collection approach and accuracy of the tool, local pilot surveys were conducted at two pharmacies in Lusaka (data not included in final analysis).

Data collection

All data were collected by a single trained investigator between February 20 and March 15, 2019. In keeping with the descriptive cross-sectional study design, data on all variables were collected at a single time point for each pharmacy. A print version of the data collection form was prepared with unique identifiers for all pharmacies and backups. Information on medicine availability and prices was recorded with the aid of skilled pharmacy staff present during the survey visits, i.e. pharmacists or pharmacy technicians. Two forms were surveyed for each medicine, namely the original brand product (OBP), more specifically the brand-name proprietary product that first obtained worldwide market authorisation, and the lowest-priced generic (LPG), defined as the cheapest generic equivalent found at each pharmacy at the time of the survey [18]. Generic insulin refers strictly speaking to biosimilar insulin [31] but was included under LPGs for reasons of simplicity. Data were only collected for specified strength(s) and dosage form(s). To be classified “available”, medicines had to be physically seen by the researcher. Price data were only recorded for medicines in stock and taken from product labels, price schedules, computer databases or, in the absence thereof, based upon verbal communication with staff. Legibility and completeness of data collection forms were verified before leaving pharmacies. Obtained data were transferred into a Microsoft Excel workbook; a double-entry technique was used to ensure accuracy of data entry. In informal discussions during the data collection, interesting and relevant information was shared by participating pharmacy staff, which was documented as field notes by the researcher. Information providing additional insights into the topic of study that was deemed to merit inclusion in the article was used to inform the discussion.

Data analysis

Analyses were carried out separately for the three provinces due to the differences in both the sample sizes and provinces’ mean per capita monthly household incomes, which were used as income measures in the affordability analysis.

Assessment of availability

Availability was assessed as the percentage of pharmacies where the medicine was found at the time of the survey and as mean percentage availability across the basket of selected antidiabetics and antihypertensives. In addition, mean percentage availability of all survey medicines was calculated for each pharmacy. Following the approach of previous studies [32-34], 5 categories were established to classify availability: absent (0%), very low (<30%), low (30-<50%), fairly high (50-<80%), high (≥80%).

Assessment of prices

Median prices were calculated in local currency (Zambia Kwacha) and converted to US$ using the Bank of Zambia’s Zambia Kwacha-US$ exchange rate [35]. The mean of the daily average exchange rates during the data collection period was used. On this basis, median price ratios (MPRs), i.e. a ratio of the local price to MSH IRPs, were calculated for medicines available at a minimum of 4 pharmacies within each province [18,29]. Interquartile ranges (IQRs) were computed to estimate the extent of price variation across pharmacies and medicines [18]. Following a 2005 WHO study on chronic disease medication, MPRs ≤ 2.5 were considered acceptable; values above this threshold were judged as excessive local prices [32]. Moreover, for medicines for which both OBPs and LPGs were found in at least 4 pharmacies per province, the potential saving for a customer through opting for the generic product was calculated as a saving ratio: Saving ratio = (price−price) / (price * 100%).

Assessment of affordability

Treatment affordability was defined as the ability to pay for a standard treatment course for diabetes or hypertension, expressed as the number of days’ income needed for its purchase [18]. A medicine was classified affordable if not exceeding a maximum of one day’ salary for a STC [18]. The daily salary was estimated from Zambia’s absolute poverty line or from the mean per capita monthly household income in the provinces [19]. Values were taken from the 2015 Living Conditions Monitoring Survey Report [19], inflated to 2018-Zambia Kwacha by use of Zambian consumer price indices (latest available data) [36,37] and converted to 2018-US$ using the same exchange rate as for the price assessments. The absolute poverty line lay at US$13.15 per month and average monthly incomes in the provinces were US$24.10 (Central), US$46.25 (Copperbelt) and US$48.85 (Lusaka). Defined daily dosages were used to approximate STCs and were taken from the WHO Collaborating Centre for Medicine Statistics Methodology [38]. For medicines available in at least 4 pharmacies per province, STCs were multiplied with median medicine prices to obtain costs of a 30-day treatment. In line with target nine in the WHO Global monitoring framework for NCDs, stipulating an 80% availability of affordable EMs in public and private facilities [39], a comprehensive analysis of both medicine availability and affordability was additionally conducted. In this way, antidiabetics and antihypertensives considered available (available in ≥80% of pharmacies surveyed) and affordable (STC ≤ one day’s salary) were identified.

Ethics statement

Before commencing data collection, the study was officially reviewed by Clinton Health Access Initiative’s internal Scientific and Ethical Review Committee (SERC). The committee deemed the study as non-human subject research as no data were collected about an individual but only on medicine availability and prices (despite the interaction with human subjects to get information during the data collection at the pharmacies following the obtainment of verbal informed consent). Consequently, no formal ethical approval was sought from the University of Zambia’s Biomedical Research Ethics Committee. Verbal informed consent was obtained from all participants included in the study. Pharmacy managers and/or staff were provided all relevant information including study background, purpose, methods, potential risks and benefits as well the voluntariness of participation before consenting to data collection. Information was provided as a written leaflet in Zambia’s official language English. The leaflet had been developed together with a Zambian pharmacist.

Results

Of the 32 selected medicines, 12 had no OBP registered with ZAMRA and were hence excluded from the survey. The survey medicines thus comprised 14 OBPs (Diabetes: 6; Hypertension: 8) and 32 LPGs (Diabetes: 7; Hypertension: 25). Data on prices and availability of these were collected from 99 pharmacies in the selected provinces Central, Copperbelt and Lusaka, including eight backups because pharmacies of the primary sample were not willing to participate (n = 2), not found in the location indicated (n = 4) or no longer in operation (n = 2). With regards to the distribution of the pharmacies in the study area, the surveyed pharmacies were invariably located in urban areas. Even in Central province, all surveyed pharmacies were located in the province’s major town with over 200,000 inhabitants.

Availability

Mean availability across all surveyed antidiabetic and antihypertensive medicines in the surveyed pharmacies was low (30-<50%) in all provinces (Table 1). Assessing availability for only medicines included in the WHO Model EML [23] showed slightly higher mean availability, yet estimates still fell into the low-level availability category. In total, 6 pharmacies had high-level stock availability (≥80%) of all surveyed medicines.

Table 1

Availability of all medicines in the surveyed pharmacies.

Province		Central (n = 6)		Copperbelt (n = 28)		Lusaka (n = 65)
Mean availability all surveyed medicines		35.14%		42.47%		42.01%
Mean availability WHO EML medicines		41.11%		46.90%		44.62%
Availability category		No.	%	No.	%	No.	%
<30%	Very low	3	50.00	5	17.86	19	29.23
30-<50%	Low	3	50.00	16	57.14	25	38.46
50-<80%	Fairly high	0	0.00	5	17.86	17	26.15
≥80%	High	0	0.00	2	7.14	4	6.15

EML = essential medicines list; n = number of surveyed pharmacies; No. = number; WHO = World Health Organization; % = percentage

Table 2 shows the availability profile of the surveyed medicines. Overall mean availability across the baskets of antidiabetics and antihypertensives was low (30-<50%) in all provinces. Regarding availability of individual medicines, high-level availability (≥80%) in all provinces was found for 2 generic antidiabetics and 9 antihypertensives. The majority of the remaining medicines was available at low (30-<50%) and very low (<30%) levels. Individual medicine availability was further found to vary with medicine forms. LPGs were more readily available than OBPs in all provinces; no single antihypertensive OBP was found in Central province. Regarding insulin in particular, all types had low or very low availability for both OBPs and biosimilars. The least available antihypertensives were injectable Hydralazine and Metoprolol (LPGs). Bearing in mind the low total sample size (n = 6) in Central province, 16.67% indicates medicines being available in one pharmacy only.

Table 2

Availability (in percentage) of surveyed antidiabetics and antihypertensives.

Province	Central		Copperbelt		Lusaka
Medicine	OBP	LPG	OBP	LPG	OBP	LPG
Glibenclamide 5mg	16.67	100	21.43	96.43	30.77	93.85
Gliclazide 80mg	-	50	-	32.14	-	38.46
Glimepiride 2mg	0	83.34	17.86	42.86	18.46	49.23
Insulin 30/70 soluble/isophane	16.67	33.34	39.29	10.71	29.23	9.23
Insulin Intermediate-acting	0	16.67	46.43	17.86	38.46	7.69
Insulin Short-acting	16.67	16.67	42.86	10.71	40	4.62
Metformin 500mg	0	100	7.14	100	9.23	100
Mean availability all surveyed antidiabetics	8.33	57.14	29.17	33.53	27.69	43.30
	34.62		36.09		37.37
Amlodipine 5mg	0	100	7.14	96.43	13.85	92.31
Amlodipine 10mg	0	83.34	7.14	89.29	6.15	89.23
Atenolol 50mg	0	100	10.71	100	21.54	93.85
Bendroflumethiazide 5mg	-	16.67	-	50	-	38.46
Bisoprolol 5mg	-	50	-	57.14	4.62*	66.15
Captopril 25mg	-	50	-	39.29	-	40
Carvedilol 6.25mg	-	33.34	7.14*	42.86	-	56.92
Carvedilol 25mg	-	0	7.14*	0	6.15*	18.46
Enalapril 5mg	-	100	-	100	-	95.38
Enalapril 10mg	-	100	-	100	-	95.38
Enalapril 20mg	-	66.67	-	100	-	83.08
Hydralazine 25mg	-	0	-	21.43	-	21.54
Hydralazine Injection	-	0	-	3.57	-	6.15
Hydrochlorothiazide 25mg	-	66.67	-	71.43	-	52.31
Hydrochlorothiazide 50mg	-	0	-	17.86	-	30.78
Lisinopril 5mg	0	0	17.86	42.86	15.38	40
Lisinopril 10mg	0	0	17.86	39.29	16.92	41.54
Lisinopril 20mg	0	0	14.29	10.71	16.92	18.46
Losartan 50mg	-	100	-	89.29	-	83.08
Metoprolol 100mg	-	0	7.14*	0	4.62*	6.15
Nifedipine 10mg	-	83.34	-	89.29	1.54*	80
Nifedipine 20mg SR	-	100	-	100	-	95.38
Propranolol 10mg	0	16.67	10.71	0	13.85	10.77
Propranolol 40mg	0	50	32.14	71.43	18.46	67.69
Verapamil 40mg	-	16.67	-	21.43	-	16.92
Mean availability all surveyed antihypertensives	0	46.67	14.73	54	15.38	53.60
	35.35		37.23		44.34

LPG = lowest-priced generic; mg = milligram; OBP = original brand product; SR = sustained release.

-: Not surveyed original brand products due to lack of official market registration in Zambia

*: Original brand product without official market registration in Zambia that was found in surveyed pharmacies.

Prices

The majority of medicines was sold at markedly higher prices than their international reference prices in all provinces (Table 3; values for Central province only available for LPGs). In total, 9 LPG antihypertensives had MPRs <1, indicating that Zambian patient prices for these are cheaper than external reference standards. An additional 4 generic diabetes and 20 hypertension medicines had acceptable prices (MPRs ≤2.5). The remainder had excessive MPRs above 2.5. Large variations in MPRs were observed across pharmacies (illustrated by the IQRs), individual medicines and between medicine forms. MPRs of OBPs were consistently, often multiple times, higher than that of their generic equivalents. The biggest differences were found for Amlodipine (5/10mg) and Glibenclamide, in which cases MPRs of OBPs were 28.39, 18.94 and 13.34 times greater than that of their generic form. Median unit prices can be found in S1 Table.

Table 3

Median price ratios (interquartile ranges) of surveyed medicines.

Province	Central		Copperbelt				Lusaka
Medicine form	LPG		OBP		LPG		OBP		LPG
	MPR	(IQR)	MPR	(IQR)	MPR	(IQR)	MPR	(IQR)	MPR	(IQR)
Glibenclamide 5mg	4.83	(4.31–6.39)	55.23	(48.22–69.04)	4.14	(3.97–5.87)	49.02	(41.42–62.14)	5.52	(4.14–6.90)
Gliclazide 80mg	-	-	-	-	2.26	(1.60–2.61)	-	-	1.62	(1.54–2.42)
Glimepiride 2mg	3.94	(2.30–3.94)	15.41	(9.76–28.94)	6.30	(4.48–7.23)	16.99	(12.74–26.32)	2.95	(2.30–3.98)
Insulin 30/70	-	-	28.82	(26.93–31.70)	-	-	21.41	(19.76–23.06)	16.80	(15.65–18.12)
Insulin Intermediate	-	-	24.79	(21.89–27.98)	11.33	(10.27–11.33)	18.42	(17.00–21.22)	11.55	(10.59–12.15)
Insulin Short-acting	-	-	22.93	(19.94–26.58)	-	-	17.28	(15.95–19.94)	-	-
Metformin 500mg	2.23	(1.64–2.62)	-	-	2.1	(1.57–2.62)	10.49	(10.49–10.49)	2.62	(1.57–2.62)
Amlodipine 5mg	1.87	(1.56–2.37)	-	-	2.49	(1.99–3.49)	47.16	(39.02–59.78)	2.49	(1.99–2.99)
Amlodipine 10mg	1.38	(1.27–2.12)	-	-	1.59	(1.06–1.70)	36.06	(31.91–36.06)	1.27	(1.06–1.70)
Atenolol 50mg	2.76	(2.30–3.49)	42.03	(39.40–42.03)	2.94	(2.21–3.68)	39.14	(29.42–42.03)	3.68	(2.63–4.41)
Bendroflum. 5mg	-	-	-	-	7.36	(6.74–7.60)	-	-	7.12	(5.89–7.60)
Bisoprolol 5mg	-	-	-	-	2.11	(1.98–2.47)	-	-	1.47	(1.25–1.87)
Captopril 25mg	-	-	-	-	1.60	(1.48–1.68)	-	-	1.60	(0.96–1.92)
Carvedilol 6.25mg	-	-	-	-	4.10	(3.56–4.71)	-	-	3.37	(2.58–3.67)
Carvedilol 25mg	-	-	-	-	-	-	19.72	(18.38–19.71)	4.83	(4.77–6.73)
Enalapril 5mg	2.84	(2.36–3.31)	-	-	3.41	(2.27–3.78)	-	-	3.78	(2.27–3.78)
Enalapril 10mg	0.85	(0.67–1.29)	-	-	0.89	(0.71–1.25)	-	-	0.89	(0.89–1.22)
Enalapril 20mg	4.14	(4.14–5.52)	-	-	4.14	(3.45–7.77)	-	-	4.14	(3.45–6.82)
Hydralazine 25mg	-	-	-	-	4.64	(2.51–5.80)	-	-	5.00	(4.51–5.80)
Hydralazine Injection	-	-	-	-	-	-	-	-	0.46	(0.20–0.71)
HCT 25mg	6.86	(4.35–11.44)	-	-	17.39	(9.15–20.13)	-	-	13.73	(9.15–18.30)
HCT 50mg	-	-	-	-	8.03	(8.03–20.08)	-	-	8.03	(8.03–16.06)
Lisinopril 5mg	-	-	6.79	(6.62–7.60)	1.83	(1.31–2.11)	5.86	(4.29–7.45)	1.41	(1.12–1.53)
Lisinopril 10mg	-	-	4.61	(4.49–5.16)	1.43	(1.14–1.55)	3.97	(3.57–5.34)	1.11	(0.95–1.40)
Lisinopril 20mg	-	-	20.57	(17.49–23.65)	-	-	15.12	(14.48–22.26)	7.88	(4.58–10.13)
Losartan 50mg	0.82	(0.80–0.97)	-	-	0.80	(0.68–0.91)	-	-	0.81	(0.58–1.03)
Metoprolol 100mg	-	-	-	-	-	-	-	-	6.56	(1.92–13.09)
Nifedipine 10mg	0.83	(0.83–0.95)	-	-	0.95	(0.71–1.19)	-	-	1.19	(0.92–1.42)
Nifedipine 20mg SR	1.86	(1.69–2.28)	-	-	1.69	(1.35–1.69)	-	-	1.86	(1.69–2.36)
Propranolol 10mg	-	-	-	-	-	-	5.39	(4.70–6.47)	2.70	(1.23–3.27)
Propranolol 40mg	-	-	45.63	(38.78–57.03)	3.99	(3.42–5.70)	35.36	(25.95–57.03)	5.70	(3.42–6.48)
Verapamil 50mg	-	-	-	-	6.28	(3.94–11.12)	-	-	2.05	(1.74–8.83)

Bendroflum. = Bendroflumethiazide; HCT = Hydrochlorothiazide; IQR = interquartile range; LPG = lowest-priced generic; mg = milligram; MPR = median price ration; OBP = original brand product; SR = sustained release.

Fig 1 illustrates these variations for medicines found as both OBPs and LPGs in at least 4 pharmacies in Lusaka province. Based on the findings, the saving ratios showed savings associated with opting for generics between 21.54% and 92.5% for antidiabetics and 47.87% to 96.47% for antihypertensives.

Fig 1

Variations in median price ratios between original brand and generic products in Lusaka province.

LPG = lowest-priced generic; mg = milligram; OBP = original brand product.

Affordability

In comparison to the national absolute poverty line, the monthly median treatment cost was considered unaffordable for the entire spectrum of surveyed medicines in all provinces. No standard treatment course with any of the medicines cost less or equal than one day’s wage (affordability threshold). If affordability instead was measured against mean per capita monthly household income, the mean affordable proportion increased from 0% to 31% in Central and to 24% in Copperbelt and Lusaka province. Treatment with a number of LPGs lay within an affordable range in each province (Table 4). By contrast, OBPs were still unaffordable at the average provincial income levels with STC expenditures well above the threshold. The same applies to all insulins, requiring 3.29 to 7.20 days’ salaries to purchase a 30-day STC.

Table 4

Lowest-priced generics considered affordable (using mean per capita income).

	Province
	Central	Copperbelt	Lusaka
National absolute poverty line level monthly income	$13.15 ($0.63/day)
Mean per capita monthly household income	$24.10 ($1.15/day)	$46.25 ($2.20/day)	$48.85 ($2.33/day)
Medicine
Glibenclamide 5mg	Not affordable (1.53)	0.68	0.86
Gliclazide 80mg	-	Not affordable (1.20)	0.81
Amlodipine 5mg	0.82	0.57	0.54
Amlodipine 10mg	0.71	0.43	0.32
Atenolol 50mg	Not affordable (1.23)	0.68	0.81
Bendroflumethiazide 5mg	-	0.79	0.72
Enalapril 10mg	Not affordable (1.04)	0.57	0.54
Enalapril 20mg	0.66	0.34	0.32
Hydrochlorothiazide 25mg	0.82	Not affordable (1.08)	0.81
Hydrochlorothiazide 50mg	-	0.28	0.27
Nifedipine 20mg SR	Not affordable (1.80)	0.85	0.81

mg = milligram; SR = sustained release; $ = US-Dollar.

Fig 2 illustrates the availability-affordability analysis in Lusaka province (using the mean per capita household income as base for affordability estimates). The x-axis represents the number of days’ wages required to purchase an STC, the y-axis shows availability levels. Plotting the defined affordability threshold and the ≥80% availability target divides the plane into 4 quadrants: north-east (high availability/low affordability), south-east (low availability/low affordability), south-west (low availability/high affordability) and north-west (high availability/high affordability). 7 LPGs were located in the north-west quadrant and met both targets, including Glibenclamide for diabetes and Amlodipine (5/10mg), Atenolol, Enalapril (10/20mg) and Nifedipine (20mg SR) for hypertension. The majority of medicines for both NCDs (n = 26) was located in the south-east quadrant displaying low availability <80% and low affordability. OBPs were without exception located in this quadrant. The same seven generics met both targets in Copperbelt; only Amlodipine (5/10mg) was found available and affordable in Central province.

Fig 2

Availability-affordability analysis of surveyed medicines in Lusaka province.

LPG = Lowest-priced generic; OBP = original brand product.

Registration status

Another finding was the registration status of the medicine names and manufacturers recorded in pharmacies. As LPGs were not prespecified, the actual products stocked varied greatly across pharmacies. Between 2 and 10 generic brands were found per medicine. Of these, 71.04% (Central), 60.85% (Copperbelt) and 54.69% (Lusaka) had ZAMRA market authorisation. No single generic product found in any pharmacy was officially registered for several LPGs, such as Bendroflumethiazide, Lisinopril (all strengths) and Verapamil.

Discussion

To our knowledge, this is the first study investigating availability, prices and affordability of antidiabetic and antihypertensive medicines in private sector pharmacies in Zambia. The results showed that the majority of surveyed medicines was inadequately available (<80%), most prices were high in an international comparison and treatment with these medicines was largely unaffordable against the set affordability thresholds.

Mean availability of all medicines across pharmacies and the two surveyed baskets of medicines was low, which reflects the inadequate availability (<80%) of the majority of individual medicines. Nevertheless, it is important not to overrate low availability as pharmacies may stock different strengths of the medicines and the applied cross-sectional design does not capture stock fluctuations over time. Individual medicine availability should further be considered in the light of availability of other medicines within the same therapeutic class that would allow for therapeutic substitution [40] (e.g. Lisinopril/Enalapril). As no substitution possibilities exist for insulin, access is vital for treatment success and ultimately life-saving [3] and its low availability hence striking. In 2005, Beran et al. found insulin available at all 13 public hospitals and 42% of health centres surveyed across three Zambian provinces [41], yet in 2016, Kalungia et al. found low-level availability of 22.2% to 37.8% at 15 surveyed public facilities in Lusaka [12]. Moreover, nearly all medicines were more widely available as generics compared to OBPs, which was reflected in the results of similar studies [7-10,33,42-49]. In contrast, the higher availability of original compared to biosimilar insulin (Copperbelt and Lusaka) was remarkable. This could be explained by the lack of patient acceptance due to perceived lower effectiveness of biosimilar insulin [50] and the limited competition on the global insulin market. Novo Nordisk, Eli Lilly and Sanofi Aventis dominate and control about 96% of the market volume and 99% of its value [51]. Further, they have direct distribution agreements with countries [52], which is reflected by the fact that insulins included in the ZEML are partly listed by Novo Nordisk’s product brand names rather than their international non-proprietary names [24].

In similar studies, with few exceptions, Metformin and Glibenclamide were also found the most widely available medicines [9,32,33,43,44,47,49]. Ewen et al. reported higher private sector availability (66%) for a nearly identical basket of antidiabetics in 12 LMICs [10], thus indicating improvement potential in the Zambian provinces. Two large-scale reviews reported estimates for generic Atenolol, Captopril, Hydrochlorothiazide, Losartan and Nifedipine [8,32]; availability of these medicines was markedly higher in the present study, which, given generics’ generally lower prices, can be evaluated positively.

The poor availability is likely to be multifactorial. Apart from those mentioned, further possible explanations might be cash-flow constraints [2], inadequate local manufacturing capacities and resulting import dependencies [1,6], poor demand forecasting due to inadequate HMIS in many surveyed pharmacies and purchasing not being guided by the ZEML. Pharmacy managers further highlighted the varying delivery lead times based on location, size and working capital of pharmacies and further the close relationship between stock and demand (they predominantly stock fast-moving products as other medicines run a high risk of expiring). Ordering specific medicines on request or receiving stock in between regularly scheduled deliveries were indicated as possibilities–actual feasibility, however, is in turn dependent on the factors influencing delivery lead times and hardly possible for smaller or remotely located pharmacies. Further research is needed to document the significance of these findings. Similarly, pharmaceutical market mechanisms and business conditions in Zambia require further investigation, not least to identify ways of increasing availability to clearly needed antidiabetics and antihypertensives.

On the whole, median price ratios of the surveyed medicines were largely found above the stipulated ≤2.5 threshold. Values of OBPs were consistently higher than of generic equivalents, which is underpinned by the high saving ratios between 21.54% and 96.47%. The results of similar surveys in LMICs corroborate these findings [8,9,46-49,10,13,32,41-45]. The variability in MPRs between individual medicines has also become apparent in similar surveys, albeit with considerably different MPRs even for the same medicines. For instance, MPRs ranged from 2.8 [42] to 44.31 [8] for Atenolol, from 0.15 [32] to 14.70 [8] for Captopril and lay between 3.42 [9] and 37.28 for Glibenclamide (LPGs) [13]. These differences are indicative of the various factors influencing MPRs, including country-specific conditions (medicine prices, procurement systems, currency strength), IRPs and the year when studies were performed.

MPRs of insulin were further noticeable, since even biosimilars were priced at 11.33 to 16.80 times their IRP. This coincides with the findings of other studies [53,54]. Within Novo Nordisk’s “Access to Insulin Commitment” initiative, it supplies insulin to LMICs’ governments at a ceiling price of US$4 (vial) [55]. This falls substantially below median prices found in the surveyed provinces (US$10.87 to US$14.42) and thus confirms the findings of 85–125% higher insulin prices from private wholesalers [41]. At present prices, yearly treatment would cost 2.04 to 2.74 times the total health expenditure per capita in Zambia of US$69.4 (2015) [56] and 9.3% to 12.3% of annual GDP per capita of US$1513.3 (2017) [57]. Pricing data reported echo the high prices found, albeit with great variations. In the WHO African region, HAI reported that prices for short-acting insulin ranged from US$3 (Senegal) to US$47 (Namibia) [58].

Several reasons might explain the high MPRs. Firstly, excessive fluctuations in foreign currency exchange rates and the depreciation of the ZMW-US$ parity have led to price instability and increases in prices for import goods [59]. As the bulk of medicines is imported and procured in US$ [1,6], it can be expected that such externalities have forced prices to go up. Further, this study focused on patient prices and procurement prices and charges (e.g. mark-ups, taxes, distribution costs) accruing throughout the supply chain were not considered. These were shown in other studies to account for price differences and increases in studies including price component analyses, with retail mark-ups being the highest contributor [8,9,13,32,34,46,48,49]. As medicines are exempted from import tariffs and VAT in Zambia [60,61], present study results similarly suggest substantial mark-ups for the surveyed medicines. Moreover, there are currently no legal or regulatory provisions regarding medicine pricing in the private sector [62]. For instance, mark-ups are neither reduced nor regulated [62] and prices thus, at best, determined by market forces. The resulting financial scope and inconsistencies in mark-ups could hence provide further explanations for the observed MPR variations. In addition, there is little competition in the supply of antihypertensive and antidiabetic medicines; the largest seller accounts for 91.7% and 100% of all sales, which has a direct bearing on procurement and patient prices of these medicines [63]. Exact causes for the high prices found, however, need to be ascertained through a price components study and additional explorative studies.

Medicine affordability is not a straightforward concept and lacks a clear definition [10,64]. However, the proposed measure in the WHO/HAI methods (number of days’ wages to purchase a standard treatment cost) has been widely recognised as it clearly and unambiguously shows the financial burden for those having to purchase their medicines individually [10,18]. In the present study, affordability of surveyed antidiabetics and antihypertensives against the defined threshold (STC ≤ one day’s wage) was very low; no medicine was considered affordable against the national absolute poverty line and only 4 (Central), 9 (Copperbelt) and 11 LPGs (Lusaka) had treatment costs below the threshold when using provinces’ mean per capita monthly household incomes as measure. Hence, these results indicate that hypertension and diabetes place a high economic burden on individuals and households at these or lower income levels.

A 2009 study found Glibenclamide, Captopril, Hydrochlorothiazide and Nifedipine affordable in the private sector in China using the Chinese absolute poverty line (US$0.4) [34] as base for defining affordability, thus contrasting the present findings. Most other relevant surveys based the affordability assessment on the salary of the lowest paid unskilled government worker, which limits comparability of results from various studies. However, as that salary generally is higher than the affordability thresholds used in this study, one can assume that medicines considered unaffordable in those other studies would also be classified as that in the present analysis. Firstly, treatment costs with OBPs in general [8,10,32-34,42,43,45,47,65] and with any original or biosimilar insulin in particular [10,33,42,43,45-49,65] were also found unaffordable against the threshold. Further coinciding with the present findings, Glibenclamide and Hydrochlorothiazide were found to be the most affordable medicines for treating diabetes and hypertension, respectively [32-34,42-45,47]. Results for the remaining medicines differed across studies. For instance, as was the case in Copperbelt and Lusaka province, Atenolol, Enalapril and Nifedipine were affordable in several studies [32,34,44] and exceeded the threshold in others [42,43,45]. This shows the high context-sensitivity of affordability analyses and the dependency of results on factors such as medicine prices and income levels.

Additional aspects need to be considered in relation to affordability. First, the absolute poverty line was used instead of the extreme (food) poverty line [19]. However, a large number of the provinces’ residents have incomes below the absolute poverty line [19], making medicines even more unaffordable for them. Income distribution should also be considered when interpreting affordability based on provincial average incomes. The gini coefficients of around 0.50 indicate high economic inequalities in all provinces [19]–even medicines considered affordable are thus still beyond the reach of many patients in these provinces. Further, despite the threshold, it is difficult to accurately assess affordability, since it strongly depends on each patient’s situation. For instance, defined daily dosages mask individual dosage requirements and intake of only one medicine was assumed in the analysis. Assessments were based on medicine costs alone and disregarded additional expenses, such as cost of supplies for medicine administration and travel costs. Including such factors could change affordability results for the worse. Moreover, the low average quantity sold by most pharmacies further emerged as crucial. Medicines were almost exclusively sold in strips (10 tablets) or as single tablets. Pharmacists’ explanation to this was that the limited financial means of many customers prevented them from buying full STCs. Instead, they need to visit the pharmacies several times a month (or even daily) in order to obtain the required medication in smaller quantities at, hence, lower costs at one time. In light of mentioned travel costs and as bulk-breaking has been shown to entail higher mark-ups and prices [2], the purchase of smaller quantities could further compromise affordability. Additionally, it illustrates the close link between affordability and availability: if patients cannot afford to build up adequate private medicine reserves, unavailability can lead to erratic treatment and people having to forego care. As with availability, the significance of this information needs to be documented by further research. Further, it is vital to pay due attention to long-term affordability, which has been shown a main cause of poor adherence to long-term treatment for NCDs and impoverishment [66-68]; future research should include such analyses.

The nature of medicines makes it impossible for patients to distinguish levels of quality and efficacy [69]. Consequently, the lack of registration detected may have serious effects on individual and public health. Quality, efficacy and safety of unregistered medicines are not adequately tested and ensured, hence increasing the possibility of substandard, falsely labelled, and falsified products entering the market [69]. WHO estimates that one in ten medicines in LMICs are falsified or substandard [70], a fact that also raises concerns for Zambia.

Study limitations

As provinces were purposively selected, the possibility of any selection bias of provinces introduced cannot be entirely ruled out. A larger study surveying the remaining 7 provinces should be conducted to allow for conclusive statements on availability, prices and affordability of antidiabetics and antihypertensives in private pharmacies in Zambia. Data on all variables were collected at a single time point for each pharmacy, which might not give a correct picture of the current or long-term availability and/or prices. Future studies should examine these variables to allow for an assessment of the commodity security of antidiabetic and antihypertensive medicines. Further, the list of surveyed medicines was not exhaustive. It did not account for alternate strengths or dosage forms and medicines found in many pharmacies (e.g. Telmisartan and fixed-dose combinations including Metformin-Glimepiride, Hydrochlorothiazide-Losartan) were excluded (no IRPs available in the MSH price set). Future research would benefit from recording availability and price information of such medicines, even if MPRs cannot be calculated. Moreover, many pharmacies had no price lists or IT-systems in place to extract medicine prices, making it necessary for the researcher to rely on information provided by pharmacy staff. To minimize bias and the possibility of incorrect information given, data were recorded with either pharmacists or pharmacy technicians and one can thus assume that valid and reliable data were obtained. Moreover, price assessments relied on median price ratios, the accuracy of which depends on the reliability of IRPs published by MSH. If these median values are based on few available data and thus potentially skewed by high/low supplier prices, this biases MPRs and makes the Zambian prices appear more/less favourable in an international comparison. Lastly, the qualitative information included in the discussion obtained in from participating pharmacy staff warrants further examination in future qualitative studies.

Conclusion

This study provided evidence on availability, prices and affordability of antidiabetic and antihypertensive medicines in private pharmacies in Central, Copperbelt and Lusaka province of Zambia. Diabetes and hypertension patients need a reliable supply of affordable medicines to avoid preventable morbidity and mortality. However, study results showed that the majority of surveyed antidiabetics and antihypertensives were insufficiently available and that they were largely unaffordable against the defined thresholds. The next steps now needed are to assess the exact causes for these results and to conduct similar surveys in the other Zambian provinces as well as in the public sector to enable conclusive statements about access to antidiabetic and antihypertensive medicines for the health sector as a whole. On this basis, policy measures and strategies should then be devised and implemented at various levels of the Zambian healthcare system and more generally the pharmaceutical market.

Median unit prices (in 2019 US-Dollars) of surveyed antidiabetics and antihypertensives.

(DOCX)

Click here for additional data file.

16 Sep 2019

PONE-D-19-20413

Availability, prices and affordability of essential medicines for treatment of diabetes and hypertension in the private sector in Zambia.

PLOS ONE

Dear Dr. Forsberg,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

We would appreciate receiving your revised manuscript by Oct 31 2019 11:59PM. When you are ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter.

To enhance the reproducibility of your results, we recommend that if applicable you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). This letter should be uploaded as separate file and labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. This file should be uploaded as separate file and labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. This file should be uploaded as separate file and labeled 'Manuscript'.

Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

We look forward to receiving your revised manuscript.

Kind regards,

Khin Thet Wai, MBBS, MPH, MA (Population & Family Planning Resear

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at http://www.journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and http://www.journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

Additional Editor Comments (if provided):

This paper used the WHO standardized methodology and could be able to bring out the policy-related evidence to improve access to antidiabetics and antihypertensive in the private sector at Zambia.

The significant difference median price ratios for LPG between the study sites requires an appropriate statistical test.

Apart from that authors should respond thoroughly to the modest request of reviewers to deal with sampling issues, some results and the discussion part to improve the writeup.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Line 31: "In line with its pledge to achieve Universal Health Coverage by 2030, the government..". The authors have to indicate the country as this is the first statement in the introduction.

Line 121-122: "To verify effectiveness of the data collection approach and accuracy of the tool, local pilot surveys were conducted at two pharmacies in Lusaka". Was this data included in the final analysis for this paper?

Line 197-199: "Of the 32 selected medicines, 12 had no OBP registered with ZAMRA. The survey drugs thus

198 comprised 14 OBPs (Diabetes: 6; Hypertension: 8) and 32 LPGs (Diabetes: 7; Hypertension:

199 25)." The number of OBP (12 + 14) does not match with LPG (32). Is there an explanation for this?

Line 240: "Table 3. Median price ratios (interquartile ranges) of surveyed drugs." Preference would be to include the results of Central province in the same table than text below it.

Line 259: "Affordability." The authors could have used the minimum wage for Zambia than the 'national absolute poverty line' as the study was conducted in major cities where majority of the individuals are employed in various sectors.

Reviewer #2: GENERAL COMMENTS:

The paper explored an important subject of availability, pricing and affordability of antidiabetic and antihypertensive medicines in the private sector in Zambia. Addressing treatment challenges for non-communicable diseases in low-middle income countries does require evidence to inform policy and practice interventions towards the universal goals of ensuring access to essential medicines. Though the paper was limited in scope and coverage in terms of study sites, a country-wide study would add value to the findings providing insight into the magnitude of the medicine availability for diabetes mellitus and hypertension treatment in Zambia’s private sector. Overall, the paper is relevant, well structured, has sufficient scholarly rigor, and has a logical flow of arguments, notwithstanding the limitations highlighted by authors.

SPECIFIC COMMENTS:

The title is a bit misleading. Reading through the paper, the scope of population coverage was rather limited. The paper only focussed on measuring availability, pricing and affordability in the private pharmacies in Zambia, which does not necessarily represent the entire scope of private sector providers of medicines. I note that the study did not cover the other private sector providers such as private hospitals and consulting room dispensaries. I assume these private players also do provide antidiabetic and antihypertensive medicines in the setting (Zambia)? Since the coverage of study sites was limited, I can suggest the authors revise and refocus their title to state “…availability, price and affordability in the private pharmacies in Zambia”

Line 4: It is good that the authors sampled the pharmacies from the list of registered (legal) outlets, however, among the 99 private pharmacies surveyed, the authors do not state the distribution of these pharmacies in the setting, whether and how many of these pharmacies were located in urban versus rural areas of Zambia? Did availability, pricing and affordability vary by location of the pharmacy in the setting?

Line 7: Regarding reasons why “international reference pricing” was used as a measure of affordability, did the authors consider difference in pricing considerations of originator brand products, local brands, versus ‘generic’ products, including cost variations between European versus Asian manufactured products on the market? Was there a local costing structure for medicines in Zambia? I will appreciate a comment on that.

Line 13: In the results section, a bit more clarity required when stating the findings. Some missing words and stylistic aspects in the opening sentence e.g. opening a sentence with a numeral?

Line 14: Authors state: “availability levels for the remainder …..”, but what is the remainder in this case? How many medicines were the remainder? The sentence is not clear. Consider using absolute measures.

Line 31: Which “government” are authors referring to? The opening paragraph (Line 31 to 35) needs to be recast and context provided.

Authors seem to be using the terms “drug” and “medicine” interchangeably? I highly doubt that the two terms are synonyms. I can suggest authors maintain to use one term consistently throughout the paper.

Line 38: Did the authors select the three provinces conveniently or purposively? This is not clear and if so, why? If not, a statement should be provided how authors ensured selection bias of provinces was avoided.

Line 98 – 100: Was a sample size formula applied to select the 99 pharmacies from the total population of private pharmacies? If so, what were the assumptions forwarded in the computation of sample size (notwithstanding the methods developed by Measure Evaluation)?

Moreover, how was the site selection (private pharmacy) done? Was due consideration given to systematically or randomly select the pharmacies or otherwise? A clear statement on this will be appreciated.

Line 103 – 104: How did the authors arrive at 10% for adjusting sample size? Any explanation or justification for this? Why 10% and not say 15% or 20%? What was the considered assumption for non-response rate?

Line 108: Among the medicines that consisted the 7 antidiabetic and 25 antihypertensive medicines, it is not clear whether authors were only interested in the single formulation medicines, or even co-formulated products?

It is a shame that if indeed the medicine was not available (perhaps stocked out) at the respective pharmacies on spot survey at time of data collection, the data collection tool would have also done well to collect data on whether the medicine was ever stocked, the number of stock-out days in a period specified, and so on. This finding could potentially inform whether there were underlying issues with medicine demand or supply challenges in Zambia. I feel readers of this paper would also be interested to know this information.

Line 145 - 147: The measure of availability that was used by Authors was rather unreliable, despite the method being widely applied by previous studies. There is a difference in ‘on the spot availability’ and ‘consistent availability’. It appears to me the authors measured ‘on the spot availability or presence’ of the medicines on spot survey. Assessing availability of medicines, in the context of pharmaceutical supply chain management principles, requires measuring commodity security i.e. consistency of availability to maintain steady flow of the product to majority consumers at all times. A medicine must be in stock in sufficient quantities to meet demand for a specified period of time consistently to date (without stock out) for it to be deemed available. Was commodity security of supply a consideration or on the spot availability of the product on the shelf?

Line 167: “STC” should be written in full when first used, then abbreviated thereafter.

Line 177: The subheading sounds repetitive of the two previous subheadings used. I feel this section from line 177 to 182 is not adding any value to the paper. If the information is needed, authors can take move the sentences to other sections.

Line 223: Am wondering if private pharmacies in the setting not having available some of the medicines surveyed is a surprising finding. Am not surprised by this finding since ideally, most if not all these antidiabetic and antihypertensive medicines are supplied on prescription-only basis to clients accessing private healthcare. In the private (for-profit) sector, commodity availability is often driven by demand (from prescriptions being received at the private pharmacy). Therefore, we wouldn’t expect a particular medicine that is not frequently prescribed to be so available in all the private pharmacies, more especially if that medicine is not on the EML or national standard treatment guideline (STG) for diabetes or hypertension. In actual fact, to be prescribed that often the medicine has to be at least the choice specified by the national treatment guidelines thereof. Did the authors consider this possible cause negatively impacting on availability in the pharmacies? So to avoid distorting the availability picture, Authors would do well to focus on those medicines that actually have marketing authorisation and feature on the National Formulary and/or STG rather than the WHO lists as reference of what is actually recommended to be prescribed for diabetes and hypertension treatment in Zambia. A discussion point perhaps?

The findings show that even some medicines without official marketing authorisation were actually selling at a price in the private pharmacies in Zambia. If some of the pharmacies were actually stocking and selling antidiabetic and antihypertensive medicines that had no official marketing authorisation from the regulator, did that imply there was illegal sale and distribution of some antidiabetic and antihypertensive medicines in private pharmacies in Zambia? Was that a weakness in the regulatory aspects? Do you want to comment on that?

Line 270: Regarding pricing affordability for generic brands, I feel the findings in table 4 would be better interpreted when the actual average price paid per unit of sale is shown (whether per pack, number of tablets, or monthly supply quoted either in local or US dollar currency equivalent) and also the affordability index (using mean per capita income) indicated in the table 4.

Line 335, 427 and elsewhere: In the discussion, Authors bring out what appears to be some qualitative findings from informant responses (e.g. from Pharmacy managers about “varying delivery times” and so on). If at all there was a qualitative component in the study, why were the qualitative findings/results? Did the data collection also involve qualitative data? Were there any qualitative findings in the study? This needs to be clarified from the onset.

Line 447 – 460: Authors bring out a number of important limitations of the study. They can also do well to mention how these limitations were addressed or can be mitigated in future studies.

Line 462 – 471: The conclusion needs to be recast. Line 462 to 464 are not adding value to the section. Rather than general statements, authors will do well to address the research objective(s) by mentioning the overarching findings of the study.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: Felix Khuluza, PhD

Reviewer #2: Yes: Aubrey Chichonyi Kalungia

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files to be viewed.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org. Please note that Supporting Information files do not need this step.

22 Oct 2019

We would like to thank the reviewers for their valuable comments and feedback. In the following, we respond to their comments. Please see the uploaded word document "Response to Reviewers", which contains the same information in tabular format for the sake of improved clarity.

Reviewer 1

31 “Republic of Zambia” has been added in LL33-34 to make it clear which country we are referring to.

121-122 A brief clarification that this data was not included in the final analysis for this paper has been added in LL153.

197-199 A brief statement has been included in LL236-237 explaining that the drugs without official ZAMRA registration were excluded from the survey.

240 We agree with the reviewer as regards the inclusion of the Central Province values in Table 3. The layout has been adjusted accordingly to allow for an additional column. Only values for LPGs have been included as no MPRs could be calculated for any OBP antihypertensive/antidiabetic medicine in Central province (none was available in at least 4 pharmacies).

259 We agree that using the minimum wage would certainly have given valuable answers as regards medicine affordability for those having incomes at or around this threshold. However, conducting the analyses with both Zambia’s absolute poverty line and the mean per capita monthly household income in the three provinces was a deliberate choice to represent both the poor and the social average. Firstly, especially since a considerable share of the provinces’ population had a disposable income below the national absolute poverty line (56.2% in Central, 30.8% in Copperbelt and 20.2% in Lusaka province – the national absolute poverty line lay at $13.15 per month at the time of the survey), using this measure seemed appropriate to represent this population share. Secondly, the average monthly incomes ($24.10 in Central, $46.25 in Copperbelt and $48.85 in Lusaka per month at the time of the survey) are considerably lower than the minimum wage mandated by law in Zambia (ZKW 1050 per month, equaling $87.76 using the exchange rate applied in the survey). This shows the discrepancy between the level of minimum salary applying to the share of the population with formal sector jobs and people’s actual income, most of which earn their incomes in the informal sector. On these grounds, we propose to keep the actual (mean) per capita household income as a second affordability measure.

Reviewer 2

Title The title and the objective have been revised accordingly.

4 A statement has been added about the urbanization status of the three chosen provinces (LL97-98). An additional sentence has been added in the results (LL242-245) to elaborate on the distribution of the pharmacies in the study area. As the pharmacies were invariably located in urban parts of Zambia, differences in availability, prices or affordability could not be assessed.

7 Using IRPs was part of the WHO/HAI methodology and the purpose is to make different studies from different countries comparable and highlight high medicine prices as a basis for potential policy interventions. While designing the methodology, consideration was given to potentially calculate such price differences. However, the actual data collection showed that the vast majority of medicines originated from Asia, particularly India, and only a few medicines, particularly antihypertensives, were imported from Europe. It hence did not seem that considering cost variations between the European/Asian manufactured products would be very informative. Furthermore, no product found in any pharmacy was manufactured in Zambia, hence making comparisons to local brands impossible. LL425-429 refer to the costing structure for medicines in Zambia. “Moreover, there are currently no legal or regulatory provisions regarding medicine pricing in the private sector. For instance, mark-ups are neither reduced nor regulated and prices thus, at best, determined by market forces. The resulting financial scope and inconsistencies in mark-ups could hence provide further explanations for the observed MPR variations.” Hence, there is no specific costing structure for medicines in the Zambian private retail pharmacy sector, but prices are determined by the pharmacists and the competition they are facing.

13 The results section has been slightly rephrased to improve the mentioned stylistic aspects.

14 This sentence has been rephrased to provide more clarity.

31 “Republic of Zambia” has been added in L34 to make it clear which country and government we are referring to.

We agree with the reviewer and all “drug” and “drugs” has been replaced with “medicine” and “medicines”, respectively in order to use one term consistently throughout the paper.

38 The section “survey scope” has been revised (LL91-110). In addition, a statement has been added to the limitations section to acknowledge the possibility of selection bias.

98-100 More detail has been provided on the pharmacy selection as well as the formulas used to determine the sample sizes (LL116-123). From the methods developed by Measure Evaluation, Formula 1 (see word document) was used to determine the sample size. Precision requirements were set at p � 0.2p at the 90% confidence level (t = 1.64), resulting in a relative error of 20% (relative variance V2 = 0.22). No data were available from previous surveys regarding sample design effects; as the ZAMRA list was not clustered, it was assumed low (f = 1.2). The anticipated proportion of pharmacies with the attribute of interest was similarly unknown; in line with the specifications of the methods used, p = 0.5 and hence q = 1–p = 0.5 were used, in which case the sample size is maximized. Inserting these values resulted in a sample size of n = 115 for each province. As the total population of pharmacies was predetermined and relatively small in both provinces (N = 149 in Lusaka and N = 36 in Copperbelt province), the resulting n was greater than 5% of the total. In such cases, sample size is to be adjusted with the Finite Population Correction factor (Formula 2 in word document). After correction, resulting sample sizes were nnew = 65 for Lusaka and nnew = 28 for Copperbelt province. In addition, a brief statement has been provided in LL123-126 on how the random selection was done.

103-104 Absent any other prior information, WHO’s Service Availability and Readiness Assessment (SARA) recommends increasing the sample size by at least 10% to take into account non-response, and 10% is similarly used as an example in the consulted sampling manual published by Measure Evaluation. On this basis and in the absence of any other information on private sector non-response rates in Zambia, 10% was chosen.

108 In principle, both single-ingredient and fixed-dose combinations were considered for inclusion in the survey. However, no fixed-dose combination product fulfilled both inclusion criteria, i.e. had an IRP registered in the MSH Drug Price Indicator Guide and was registered with ZAMRA. Hence, only single formulation medicines were included in the final survey list. The lack of fixed-dose combination products in the survey list is further acknowledged in the study limitations. A statement has been added in LL156-157 (data collection) to alert the reader that data was only collected at a single time point for each pharmacy. This approach was chosen in keeping with the applied WHO/HAI methodology. In addition, while the suggested data collection approach would have been preferable, most pharmacies did not have a reliable information system keeping track of stock-out days, etc., hence rendering this approach very difficult and/or impossible. In addition, the pitfalls of the chosen data collection approach are further acknowledged in the study limitations (LL 502-506).

145-147 This comment seems directly related to the previous comment. As this study employed a descriptive, cross-sectional study design, commodity security was not considered but the focus was on actual availability on the shelf at the time of the survey. More information on availability is also provided in the reply to the comment on L223.

167 This abbreviation was introduced in L151 (Section: Selection of medicines).

177 The methods section on affordability has been merged with the availability/affordability section and the subheading for the latter was taken out (LL214-221). The same has been done in the results section (LL317-319).

223 We agree with the reviewer that the finding is not very surprising- despite the very low availability of many medicines being striking. We have addressed the reviewer’s comment in the discussion in the for instance the following (LL382ff): “Pharmacy managers further highlighted the varying delivery lead times based on location, size and working capital of pharmacies and further the close relationship between stock and demand (they predominantly stock fast-moving products as other drugs run a high risk of expiring). Ordering specific drugs on request or receiving stock in between regularly scheduled deliveries were indicated as possibilities – actual feasibility, however, is in turn dependent on the factors influencing delivery lead times and hardly possible for smaller or remotely located pharmacies.” In addition, as detailed in the medicine selection paragraph in the methods section, we briefly mentioned that the Zambia National Formulary and standard diabetes and hypertension treatment guidelines were consulted for the medicine selection in order not to miss medicines that are frequently prescribed but not included on either the WHO EML or the ZEML. The findings that unregistered medicines were sold in the surveyed pharmacies implies illegal sale and distribution to some extent. Pharmacies are not allowed to bulk purchase unregistered drugs and/or to regularly sell them. However, they are allowed to import a certain quantity of those drugs on request by specific patients. However, as many unregistered products were regularly found in many of the surveyed pharmacies and sometimes even the only products available, the assumption can be made that illegal sale and distribution was happening and that there are major regulatory weaknesses.

270 A table illustrating median unit prices for all antidiabetics and antihypertensives for all three provinces has been uploaded as a supplement and this has been pointed out to the reader (LL284-285). Moreover, two additional rows have been added to show both the level of the absolute national poverty line and the mean per capita household incomes per month and per day (in 2019 USD).

335, 427, etc. The data collection did not involve qualitative data collection per se, yet several pharmacy staff highlighted the included issues, which were recorded in the researcher’s field notes. The obtained information was verified in additional discussions with practicing pharmacists as well as officials at Medical Stores Limited and on this basis, we decided to include the information to inform the discussion section of the present paper. A brief statement has been added in the data collection section (LL172-175) with regards to the qualitative information included in the discussion. In addition, the importance of documenting the significance of this information by conducting future qualitative studies has been highlighted in the discussion (LL388-389; LL485-486). We have also included a statement in the limitations section to point out to the reader that the qualitative information warrants careful interpretation (LL521-522).

447-460 We agree with the reviewer that elaborating more on the limitations and highlighting possibilities to mitigate these for future studies is important. The limitation section has been rephrased accordingly and recommendations for future studies have been provided.

462-471 The conclusion has been rephrased and overarching findings of the study have been mentioned.

Submitted filename: Response to Reviewers.doc

Click here for additional data file.

21 Nov 2019

Availability, prices and affordability of essential medicines for treatment of diabetes and hypertension in private pharmacies in Zambia.

PONE-D-19-20413R1

Dear Dr. Forsberg,

We are pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it complies with all outstanding technical requirements.

Within one week, you will receive an e-mail containing information on the amendments required prior to publication. When all required modifications have been addressed, you will receive a formal acceptance letter and your manuscript will proceed to our production department and be scheduled for publication.

Shortly after the formal acceptance letter is sent, an invoice for payment will follow. To ensure an efficient production and billing process, please log into Editorial Manager at https://www.editorialmanager.com/pone/, click the "Update My Information" link at the top of the page, and update your user information. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to enable them to help maximize its impact. If they will be preparing press materials for this manuscript, you must inform our press team as soon as possible and no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

With kind regards,

Khin Thet Wai, MBBS, MPH, MA (Population & Family Planning Resear

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

All comments of reviewers are fully addressed.

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: The authors have addressed all the comments. In addition, the authors have provided explanations where changes to my suggestions could not be taken on board.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: Dr. Felix Khuluza

6 Dec 2019

PONE-D-19-20413R1

Availability, prices and affordability of essential medicines for treatment of diabetes and hypertension in private pharmacies in Zambia.

Dear Dr. Forsberg:

I am pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please notify them about your upcoming paper at this point, to enable them to help maximize its impact. If they will be preparing press materials for this manuscript, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

For any other questions or concerns, please email plosone@plos.org.

Thank you for submitting your work to PLOS ONE.

With kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Khin Thet Wai

Academic Editor

PLOS ONE