Quality Assessment of Selected Essential Antimicrobial Drugs from Drug Retail Outlets of Selected Cities in Eastern Ethiopia

ABSTRACT.

The prevalence of substandard and falsified (SF) antimicrobial drugs is increasing around the globe. This poses a great concern for the healthcare system. The consumption of SF antimicrobial drugs has the potential to result in treatment failure, emergence and development of antimicrobial resistance, and ultimately a rise in mortality rate. The objective of this study was to assess the quality of four commonly used antimicrobials marketed in the cities of Dire Dawa and Jijiga and the town of Togo-Wuchale, which have high potential for illegal drug trade activities in Ethiopia because they are located near the border with Somalia. A total of 54 brands/samples of amoxicillin, amoxicillin/clavulanic acid, ciprofloxacin, and norfloxacin formulations were collected covertly from 43 facilities using a convenience sampling strategy from March 16 to March 29, 2022. The samples were first screened using Global Pharma Health Fund (GPHF)-Minilab protocols and then analyzed using U.S. Pharmacopoeial and British Pharmacopoeia official methods. The quality evaluation detected no falsified product; however, it showed that 14.3% of the samples failed the GPHF-Minilab screening test semiquantitatively. Overall, 22.2% of the products analyzed did not meet any of pharmacopoeial specifications assessed: 13%, 12.2%, and 11.1% of the products failed in assay, dissolution, and weight variation, respectively. Additionally, 56.3% of amoxicillin samples, 60% of amoxicillin/clavulanate, 20% of ciprofloxacin, and 54.5% of norfloxacin samples were found to be pharmaceutically nonequivalent with their respective comparator products regarding dissolution profiles. The study showed the presence of substandard antimicrobial medicines in the eastern Ethiopian market.

INTRODUCTION

Access to quality-assured essential medicines is one of the important factors in the achievement of sustainable development goals. Therefore, the quality of medicines should be consistently assured so that they can be used safely.¹ Although many millions of people still lack access to essential medicines,² the global pharmaceutical trade has grown dramatically in recent years. This growth has opened the door not only to high-quality, safe, and effective drugs but also to low-quality, potentially toxic medicines.³ The consumption of substandard and falsified (SF) medicines has the potential to result in treatment failure, development of antimicrobial resistance, increased incidence of adverse drug reactions and side effects, and increased patient mortality.⁴ Furthermore, antimicrobial-resistant pathogens are causing 700,000 deaths globally every year,⁵ with South Asia and Africa bearing the most burden.⁶ However, the epidemiology of poor-quality antimicrobials is still poorly understood and underresearched, despite being one of the contributors to antimicrobial resistance.⁷

The WHO adopted three terms to express poor-quality medicines to remove confusion and achieve a shared understanding among all levels of stakeholders. This also helps for the effective prevention of market entry of these products. Medical items that are “substandard” are approved medical products that do not satisfy their quality standards. “Unregistered” products are those that have not been approved for the markets in which they are distributed or used by the medicine regulatory authority of the country. “Falsified” drug products are those that falsely represent their identity, composition, or source.⁴

Substandard and falsified essential drug products are widely distributed all over the world, and the WHO estimates that more than 10% of medicines in low- and middle-income countries (LMICs) are either substandard or falsified. Due to the high demand and low availability of medicines, in addition to the low regulatory capabilities in LMICs, SF medicines are a cause of concern in these nations.⁸ Ozawa et al. recently reported a 13.6% prevalence of SF drug products in LMICs and 18.7% in Africa.⁹ Furthermore, incidents of pharmaceutical crime have recently been increasing at a high rate, posing a great danger to public health.¹⁰ Essential antimicrobial drugs are among the major classes of medicines that are commonly reported to be substandard and/or falsified.^11,12

A few studies have been conducted on the quality of medicines in Ethiopia, revealing failure rates that vary depending on the type of medicine, the manufacturer, and the place of origin. Antibacterials, antimalarials, and anti-TB drugs are of greatest concern, according to the studies.^13–15 One study on 196 medicine samples of various therapeutic categories from 15 developing countries, particularly those with sub-Saharan tropical environmental conditions, including Ethiopia, reported that 67% of the samples from Ethiopia were substandard, whereas 37% of the products were illegally imported and unregistered.¹³ The U.S. Pharmacopoeial Convention (USP) publishes the outcomes of data collected from medicine quality monitoring (MQM) activities in various countries throughout the world in a publicly accessible MQM database. The database shows that samples of several solid-dose antimalarial treatments (chloroquine and quinine) from Ethiopia failed quality tests between 2013 and 2017.¹⁴ Also, based on analysis of the data base, Hajjou et al. reported that 20.4% of the 4,473 antimicrobial samples examined between 2003 and 2013 from five African countries, including Ethiopia, were of poor quality. Further, 1.2% of amoxicillin samples tested throughout that period were of low quality.¹⁵

Despite the reported high prevalence of SF antimicrobial agents globally,¹⁶ in Africa and sub-Saharan Africa,⁹ little is known about the epidemiology in terms of antimicrobial medicine quality in Eastern Ethiopia; a porous border with Somalia (stateless and poor regulatory organization for many years) is one major corridor for illicit trade of goods, including medicines. Therefore, this study was intended to assess the physicochemical quality of four of the most commonly used antibacterial medicines—amoxicillin, combined amoxicillin and clavulanate, ciprofloxacin, and norfloxacin oral solid dosage formulations—circulating in health facilities and drug retail outlets within the market of selected sites in eastern Ethiopia.

MATERIALS AND METHODS

Study design.

This study was designed according to the WHO guideline for conducting quality surveys of medicines.¹⁷

Study area.

The study area is mapped based on the potential for drug product smuggling and the potential for illegal border trade at the sites. Some of the sites also had adverse climatic conditions (i.e., hot and dry), with many pharmacies/retail outlets not being properly air-conditioned, which affects the stability of medicines. The geographic locations in eastern Ethiopia, Dire Dawa city administration, and the Somali region that are considered major gateways for the unregulated entry of medicines into the country are prioritized. Within each region, the capital cities, Dire Dawa and Jijiga, were sampled because they are centers for drug distribution and could be representative of the entire region. In addition, the border town of Togo-Wuchale with the neighboring country (Somaliland/Republic of Somalia) was considered because of its porosity and hence its potential for illegal drug trade. A map of the study area showing sampling locations is shown in Figure 1.

Figure 1.

Map of the study area in eastern Ethiopia showing sampling sites.

Selection of antimicrobial agents.

The drugs were selected based on their relevance to public health, the likelihood of finding SF products, and the risk of antimicrobial resistance associated with the products. The antibacterial drugs included in this study are also in the list of essential drugs for Ethiopia.¹⁸ They were selected based on the reported highest consumption levels countrywide¹⁹ and the highest risk of SF.²⁰ The four most commonly used antimicrobial drugs—amoxicillin combined with clavulanic acid, ciprofloxacin, and norfloxacin—were selected for this study. Additionally, solid dosage formulations of the chosen antimicrobials for adults were examined because the highest reported frequency of SF antimicrobials is linked to this dosage form.²⁰

Selection of sample collection sites.

A convenience sampling technique was used for sample collection. Sampling was performed from both the public (government) and private sectors including public and private health facilities, private pharmacies, drug stores, and rural drug vendors. Public health institutions that serve larger populations received more attention for sampling. Private facilities around those health facilities and those near markets, where illegal drugs are likely to be handled, were sampled. On this basis, samples were collected from 39 private facilities and four public hospitals.

Sample collection technique.

The covert sampling (“mystery shopping”) technique was used in privately owned facilities throughout the sample collection process. In covert sampling, a mystery shopper (the first author who had trained on the method) introduced himself as a practicing clinician owning a small clinic in small towns near the study areas and asked for all available brands of the target medicines and their prices. Then, when the medicines were brought to him, the maximum variation of manufacturers of the medicines was considered to select medicines of different brands. However, this was not possible in public (government) establishments where antimicrobials are not dispensed without prescription; overt sampling was the only option and was used to collect samples. To maintain the quality of the samples and avoid quality deterioration before testing, dosage units were not taken out of the original primary or secondary packaging. The collected samples were stored as indicated on their labels.

Sample collection from the selected facilities.

From all the products (brands or generics) available per facility, one randomly selected product was purchased covertly from the private facility. All brands of each product available were collected in public health facilities where overt sampling was used, using a letter of cooperation obtained from the Department of Pharmaceutical Chemistry and Pharmacognosy, School Pharmacy, College of Health Sciences, Addis Ababa University. Emphasis was given to different batch numbers, manufacturers, and countries of origin when moving from facility to facility to represent the quality of different products that are likely to be used by patients. In addition, unregistered products were given special attention. Each sample collected was provided with a code to ensure traceability, and for each sample, the sample collection form was completed.

Ethical approval.

Ethical approval for this study was obtained from the Ethical Review Board of the School of Pharmacy, College of Health Sciences, Addis Ababa University (reference number ERB/SOP/360/2021).

Drug quality analysis.

Visual inspection.

A visual inspection tool for packaging and labeling of pharmaceutical formulations was used to examine the physical qualities of the dosage form, packaging, and labeling information. This tool was adapted from the WHO, European Union, and USP checklists (Supplemental Table 1).

Screening with GPHF-Minilab.

The samples for amoxicillin and its combination with clavulanic acid and for ciprofloxacin were screened for identity and semiquantitative tests by thin layer chromatography (TLC) analysis per the Global Pharma Health Fund (GPHF)-Minilab protocol outlined for each drug.²¹ The authentic reference products, amoxicillin 500-mg reference tablets (lot no: 788230, GPHF, Giessen, Germany), clavulanic acid/amoxicillin 125/500-mg reference tablets (lot no: JY2838, GPHF), reference tablets containing an equivalent of 250 mg of ciprofloxacin-free base (lot no: 80807V, GPHF), and other chemicals and reagents supplied with the GPHF-Minilab kit were used for screening tests. The examined samples’ thin layer chromatograms were routinely recorded using cameras. The samples of norfloxacin were not screened at this stage because the GPHF-Minilab protocol did not have a method for this drug.

Tests for hardness, thickness, diameter, and disintegration were also conducted according to the USP 2021 (USP 43) general chapters.²² ERWEKA (Heusenstamm, Germany) disintegration test apparatus (type ZT 304), and an ERWEKA hardness, thickness, and diameter tester (type TBH220) were used.

Pharmacopoeial tests were performed at the national medicine quality control laboratory at the headquarters of the Ethiopian Food and Drug Authority (EFDA) following the USP 2021 (USP 43) and BP 2021 procedures. All samples of amoxicillin and its combination with clavulanic acid and ciprofloxacin were analyzed against the respective monographs in the USP,²² whereas the norfloxacin samples were assessed against the BP monograph.²³ The reference standards, USP amoxicillin trihydrate (lot: R106H0 and LOK359, Spain and lot: R05170, Darmstad, Germany), USP ciprofloxacin (lot: R12590, Darmstad, Germany), USP clavulanic acid (lot: R071W0, Bangalore, India) and USP norfloxacin (lot: R061R0, Bangalore, India), and USP ciprofloxacin ethylenediamine analog (lot: R013T0, Bangalore, India) were used for pharmacopoeial tests. HPLC grade acetonitrile (Fisher Scientific, Loughborough, Leicestershire, UK) and methanol (Fisher Scientific, Loughborough, Leicestershire, UK), phosphoric acid (Sigma-Aldrich, Buchs, Switzerland), hydrochloric acid (37%, ACS, International Organization for Standardization reagent grade, Scharlab, Barcelona, Spain), reagent grade glacial acetic acid (Sigma-Aldrich, Germany), potassium hydroxide pellets (Blulux Laboratories, Faridabad, India), sodium hydroxide pellets (Oxford Laboratory Fine Chem LLP, Navghar, India), monobasic sodium phosphate (Sigma-Aldrich, Germany), potassium dihydrogen orthophosphate (Park Scientific limited, Northampton UK), and triethylamine (VWR chemicals, Fontenay-sous-Bois cedex, France) were reagents and solvents used in the analysis.

Identification tests using high-performance liquid chromatography (HPLC) were carried out by comparing the retention times of the sample and the reference standard as carried out in the assay procedures. Assay tests were carried out using the chromatographic conditions and solvent systems specified in the monographs and an HPLC system equipped with a UV detector (model SPD-20A, Shimadzu, Kyoto, Japan). Analytical stainless-steel columns of Waters (Wexford, Ireland) Spherisorb ODS2 C-18 column (5 μm, 4.6 × 250 mm) for analysis of amoxicillin and its combination with clavulanate; Agilent’s (Santa Clara, California) C-18 column (5 μm, 4.6 × 250 mm) for analysis of ciprofloxacin; and ACE (Aberdeen, Scotland) C-18 (5 µm, 3.9 × 150 mm) column for analysis of norfloxacin were used for HPLC analysis.

In vitro dissolution tests were carried out using an ERWEKA dissolution test apparatus (type DT 820), and the samples were analyzed using a UV-visible spectrophotometer (model UV-1900, Shimadzu) at the specified wavelengths in the monograph for amoxicillin (272 nm), ciprofloxacin (276 nm), and norfloxacin (313 nm). Amoxicillin/clavulanate samples were quantified using HPLC systems, and columns described earlier for its assay procedure were used. The 5 mL of samples withdrawn at 5, 15, 30, 45, 60, and 75 minutes for amoxicillin; at 5, 10, 15, 20, 30, 45, and 60 minutes for amoxicillin/clavulanate; at 5, 10, 15, 20, 30, and 45 minutes for ciprofloxacin; and at 5, 10, 15, 20, 30, 35, and 45 minutes for norfloxacin were replaced with an equal volume of solvents. Calibration standards prepared at concentration levels of 112, 140, 168, 196, 224, and 252 µg/mL of amoxicillin were used to generate calibration equation for the determination of the amount dissolved at each time point of amoxicillin capsule samples. Calibration curve equations generated using standards at 100, 200, 300, 400, 500, 600, 700, and 800 µg/mL of amoxicillin and 30, 60, 90, 120, 150, 180, and 210 µg/mL of clavulanic acid were used for quantification of each active ingredient in amoxicillin/clavulanate. Standards prepared at 2, 3, 4, 5, and 6 µg/mL of ciprofloxacin and 3.2, 6.4, 9.6, 12.8, 16, 19.6, and 22.4 µg/mL of norfloxacin were used to generate calibration equations and to determine the amount of drug dissolved at each time point of the respective drugs. The mean cumulative percentage of drug dissolved was calculated using the labeled amount of the drug, and all observed points were used for comparison of dissolution profiles with the respective comparator products using similarity and difference factors.²⁴

Different kinetic models were also fitted into the dissolution data of the comparator and samples in the current study to further explain the overall release of the drug from the dosage forms. The dissolution data were analyzed using DDSolver^® software, and the release mechanisms were established through model fitting using various built-in models.²⁵ Zero- and first-order kinetics, Higuchi, Hixson–Crowell, Makoid–Banakar, and Weibull were among the models used to fit the dissolution data. The mean cumulative dissolution data at each sampling point of each sample and reference product were used for model fitting.

Acceptance criteria for the quality medicines.

Following the guidelines in the GPHF Minilab manual, results of the TLC testing were categorized as pass/suspect.²⁰ The requirement for hardness was no less than 40 N of breaking strength.²⁶ The U.S. Food and Drug Administration (FDA) recommendations were used for the evaluation of thickness and diameter results.²⁷ The USP limit for acceptance value of not more than L1% or 15% for 10 dosage units was used for weight variation. The USP 2021 and BP 2021 monograph limits for compliance were used for disintegration time (30 minutes), identification, assay, and dissolution. System suitability tests were performed before each analysis. For in-vitro dissolution profile comparison, a difference factor (f₁) value of 0 to 15 and a similarity factor (f₂) value of 50 to 100 are used to decide the interchangeability of multisource products by comparing the dissolution profiles per FDA and European Medicines Agency guidelines. The difference factor (f₁) is defined by the U.S. FDA in terms of the observed percentage difference between two profiles at each time point. This factor essentially measures the relative error between the two profiles and is calculated as $f_1= \left\{\left[{\displaystyle {\Sigma }_{t=1^n}|R_t-T_t|}\right]/\left[{\displaystyle {\Sigma }_{t=1^n}{R}_t}\right]*100\right\}$. The similarity factor is a measurement of the similarity in the percentage dissolution between two profiles. This factor is just the logarithmic reciprocal square root transformation of the squared error sum, calculated by the formula $f_2=50\times \log \left\{{\left[[1+(1/n){\displaystyle {\Sigma }_{t=1^n}{\left(R_t-T_t\right)}^2}]\right]}^{-0.5}\times 100\right\}$,²⁸ where n is the number of sampling time points, and R_t and T_t are the amount of drug dissolved at time t of reference and test, respectively. The acceptance criterion for two profiles not to be distinctly different is when f₁ value is between 0 and 15. The acceptance criterion for two profiles to be markedly similar is when 100 ≤ f₂ ≥50.

For dissolution data modeling, the most popular criteria—the Akaike information criterion (AIC), the coefficient of determination (R²), and the model selection criterion (MSC)—were deployed. The goodness of fit criteria of lowest AIC, highest R², and highest MSC were applied to select the best-fit model for the release of a drug substance from the dosage form.²⁹

RESULTS

Sample information.

A total of 54 samples of the four target medicines in solid oral dosage formulations were purchased from four public hospitals and 39 private establishments (private hospitals, pharmacies, drug stores/shops, and rural drug vendors) within the study areas. The comparator products, CiproBay 500 mg and Augmentin 625 mg, were purchased from Addis Ababa, the capital city of Ethiopia. Two products from the sample, Amoxapen (sample ACD015) and Trizolin (sample NTD052), were selected as comparator products for respective amoxicillin and norfloxacin samples following the WHO recommendation for the selection of comparator products due to the unavailability of reference-listed products in the local market.³⁰ The total number and type of sampling facilities in each city are presented in Figure 2A, and the number of each product category collected in each sampling city is presented in Figure 2B. Figure 2C shows that 16.7% were locally manufactured products and 83.3% were imported products, of which 33.3% were from China, 29.6% from India, 9.2% from Cyprus, 5.5% from Turkey, 1.8% from Kenya, and 1.8% from Austria. Of these, 1.8% did not specify their country of origin and storage condition. Detailed information on the samples collected is presented in the Supplemental Table 2.

Figure 2.

(A) Distribution of sampling facility types in each study area (n = 43). (B) Distribution of samples collected in each city (n = 54). (C) Stated country of origin.

Visual inspection revealed that all the samples met the labeling and physical appearance requirements, except for one sample of amoxicillin 500-mg capsule (sample ACW010), which was did not have a full address of the manufacturer. In addition, the label did not contain information regarding storage conditions. However, this sample (ACW010) met all the tested quality parameters. Eight of the 54 samples (14.8%) were not registered by the EFDA to be marketed in the country.

All the samples screened with the TLC using the GPHF-Minilab were positively identified for their respective active pharmaceutical ingredients (APIs). However, six of the samples failed the semiquantitative screening according to the methods of the GPHF-Minilab requirement because visual comparative observation of the formed spots was out of the 80% to 120% API content range. Five of these were nonregistered amoxicillin capsule products. Representative TLC chromatograms of the analyzed samples are shown in Supplemental Figure 1. GPHF-Minilab TLC identification was complemented with the pharmacopoeial HPLC identification. All samples analyzed were observed to contain the labeled active ingredient(s), and hence no product without the stated API(s) was found. Typical HPLC chromatograms from each of the analyzed products are shown in Supplemental Figure 2.

The uniformity of dosage units was assessed using the USP 2021 weight variation test procedure.²² The procedure was applied to 10 dosage units of each sample by using the mean weight and assay value for each sample. Then the API content in each dosage unit and acceptance values were calculated. The weight variation test showed that six of the samples (five amoxicillin and one of its combinations with clavulanate) did not meet the first stage of the USP requirement for uniformity of dosage units. The results of weight variation tests are presented in Tables 1 and 2.

Table 1.

Summary of the results of quality parameters of amoxicillin 500-mg capsules

Sample code	VI	GPHF-Minilab screening		HPLC I‡	DT (min) n = 6	Weight variation		Mean % content (± % RSD)	PDR (±% RSD) (n = 6)
		I*	Q†			Mean weight (mg)	AV
ACJ001	Pass	Pass	Pass	Pass	9.3	607.9 (5.2)	2.3	98.3 (0.4)	92.7 (2.4)
ACJ002	Pass	Pass	Pass	Pass	7.3	591.5 (5.9)	3.5	97.3 (0.5)	107.3 (1.2)
ACJ003	Pass	Pass	Pass	Pass	9.9	591.0 (4.8)	2.9	97.6 (0.5)	109 (1.2)
ACJ004	Pass	Pass	Pass	Pass	6.7	604.7 (16.5)	9	95.8 (1.5)	98.7 (2.4)
ACJ005	Pass	Pass	Pass	Pass	3.5	593.5 (7)	3.5	97.7 (0.9)	65.5 (0.4)
ACJ006	Pass	Pass	Pass	Pass	6.8	593.5 (7)	14	112.3 (1.1)	96.9 (2)
ACJ007	Pass	Pass	Suspect	Pass	8.5	568.2 (22.2)	39.1	65.5 (1.6)	–
ACJ008	Pass	Pass	Suspect	Pass	8.5	560.2 (23)	44.4	60.0 (1.9)	–
ACJ009	Pass	Pass	Pass	Pass	8.7	593.1 (11.4)	6.6	103.3 (0.4)	107.6 (2.4)
ACW010	Pass	Pass	Pass	Pass	11.1	585 (13.4)	6.4	102.2 (0.5)	103.5 (2)
ACW011	Pass	Pass	Suspect	Pass	7.5	555 (19.1)	41	62.7 (1.9)	–
ACW012	Pass	Pass	Pass	Pass	11.2	603.5 (34.3)	15.1	96.5 (0.4)	107.2 (1.2)
ACW013	Pass	Pass	Suspect	Pass	7.3	572.9 (22.1)	43.6	60.6 (2.9)	–
ACD014	Pass	Pass	Pass	Pass	11.2	607.1 (2.8)	9.2	109.5 (0.6)	108.2 (0.9)
ACD015	Pass	Pass	Pass	Pass	8.5	582.9 (5)	5.4	104.8 (1.1)	106.8 (1.7)
ACD016	Pass	Pass	Pass	Pass	6.2	593.1 (4.3)	1.8	98.4 (0.2)	118.3 (1.1)
ACD017	Pass	Pass	Suspect	Pass	8.1	541.4 (39.5)	50.1	58.7 (2.6)	–
ACD018	Pass	Pass	Pass	Pass	8.4	598.3 (5.2)	2.1	101.5 (1.7)	102.5 (1.4)
ACD019	Pass	Pass	Pass	Pass	8.2	600.3 (5.3)	2.1	99.8 (0.2)	115.7 (1.4)
ACD020	Pass	Pass	Pass	Pass	7.2	581.1 (11.9)	4.9	100 (1.8)	115.2 (1.3)
ACD021	Pass	Pass	Pass	Pass	7	597.5 (7.2)	2.9	98.8 (0.2)	112.6 (2.4)
ACD022	Pass	Pass	Pass	Pass	7	586.7 (11.2)	4.6	100.7 (0.1)	114.9 (4.2)

AV = acceptance value; DT = Disintegration time; GPHF = Global Pharma Health Fund; I = identification; PDR = percent drug release; Q = semiquantitation; VI = visual inspection.

^*Thin layer chromatography identification.

^†Thin layer chromatography semiquantitative.

^‡High-performance liquid chromatography identification.

Table 2.

Summary of the results of quality parameters of amoxicillin/clavulanate 625 mg tablets, ciprofloxacin 500 mg, and norfloxacin 400 mg tablets

Sample code	VI	GPHF-Minilab screening		HPLC I	Hardness (N), n = 10	Thickness (mm), n = 10	Diameter (mm) (± SD), n = 10	DT (min), n = 6	Weight variation		Mean % content (± RSD)		Mean PDR (± RSD), (n = 6)
		I	Q						Mean weight	AV
Amox + clav 625 mg											Amox	Clav	Amox	Clav
ACTJ023	Pass	Pass	Pass	Pass	381.2 (15.4)	7.1 (0.0)	21.7 (0.0)	8.6	1120.2 (9.1)	1.9	98.8 (0.7)	107 (0.4)	105.6 (1.7)	98.7 (1.5)
ACTJ024	Pass	Pass	Pass	Pass	232.1 (17.6)	7 (0.0)	20.2 (0.0)	7.5	1018.9 (10.2)	11.4	89.3 (0.4)	110.5 (0.4)	91.6 (2.1)	69.5 (2.5)
ACTW025	Pass	Pass	Pass	Pass	328.1 (9.1)	6.1 (0.0)	20.5 (0.0)	10.2	1036.8 (17.7)	12.8	109.8 (0.5)	44.7 (0.4)	91.9 (1.2)	75.6 (3.4)
ACTD026	Pass	Pass	Suspect	Pass	383.5 (20.3)	6.3 (0.0)	19.4 (0.0)	13.2	1094.2 (12.5)	15	113.4 (2.0)	113.1 (1.6)	96 (1.2)	97.9 (0.8)
ACTD027	Pass	Pass	Pass	Pass	180.7 (13.3)	5.5 (0.0)	21.1 (0.0)	9.7	966.4 (12.4)	10.7	108.8 (0.4)	116.4 (0.4)	111.3 (3.5)	108.9 (4)
Augm.	Pass	Pass	Pass	Pass	391.1 (9.7)	6.9 (0.0)	20.2 (0.0)	9.1			96.6 (0.4)	107.9 (0.4)	94.4 (3.3)	107.7 (3)
Ciprofloxacin 500 mg											Cipro		Cipro
CTJ028	Pass	Pass	Pass	Pass	193.8 (18.0)	17.4 (0.0)	4.9 (0.0)	1.1	716 (1.9)	4.5	101.0 (0.1)		112.9 (2.5)
CTJ029	Pass	Pass	Pass	Pass	132.5 (8.6)	15.5 (0.0)	6.0 (0.0)	12	686.2 (1.5)	8.6	106.5 (0.2)		104.9 (0.5)
CTJ030	Pass	Pass	Pass	Pass	135.7 (7.7)	18.1 (0.0)	5.1 (0.1)	17.5	787.2 (1.1)	5.7	104.5 (0.3)		71.4 (3.5)
CTJ031	Pass	Pass	Pass	Pass	151.4 (9.9)	16.6 (0.0)	4.9 (0.1)	5.7	634.7 (0.7)	2.6	97.5 (0.8)		60.4 (26.4)
CTJ032	Pass	Pass	Pass	Pass	177.9 (9.8)	18.1 (0.0)	5.4 (0.0)	3.15	772.7 (1.2)	6.4	105.0 (1.3)		103.9 (5.3)
CTJ033	Pass	Pass	Pass	Pass	227.3 (19.7)	17.2 (0.0)	5.4 (0.1)	11.3	736.6 (1.5)	6.6	104.5 (0.3)		100.5 (0.6)
CTJ034	Pass	Pass	Pass	Pass	115.4 (11.6)	18.1 (0.0)	5.1 (0.0)	1	798.2 (1.2)	2.8	100.4 (3.3)		100 (2.3)
CTJ035	Pass	Pass	Pass	Pass	150.3 (6.9)	18.2 (0.0)	5.4 (0.1)	1.5	770.5 (0.5)	8.1	91.6 (3.6)		110.5 (0.8)
CTD036	Pass	Pass	Pass	Pass	193 (25.5)	17.3 (0.0)	5.4 (0.1)	12.6	741.7 (1.4)	8.2	106.3 (0.7)		102.3 (1.8)
CTD037	Pass	Pass	Pass	Pass	189 (11.3)	17.2 (0.0)	5.4 (0.0)	1.8	741.0 (0.0)	4.8	104.8 (1.2)		99.5 (1.6)
CTD038	Pass	Pass	Pass	Pass	157.0 (13.6)	15.5 (0.0)	5.9 (0.0)	2.7	686.2 (1.1)	10.9	109.9 (1.5)		104.4 (0.7)
CTD039	Pass	Pass	Pass	Pass	162.8 (26.3)	12.8 (0.0)	5.2 (0.0)	3.0	643.9 (1.5)	5.5	103.3 (0.9)		104.1 (1.6)
CTD040	Pass	Pass	Pass	Pass	139.6 (10.5)	16.5 (0.1)	5 (0.0)	7.8	637.8 (0.8)	4.2	103.9 (0.9)		83 (9.4)
CTD041	Pass	Pass	Pass	Pass	172.9 (4.1)	18.1 (0.0)	5.4 (0.0)	4.4	776.5 (0.7)	5	104.7 (1.8)		105.1 (0.7)
CTD042	Pass	Pass	Pass	Pass	168.6 (6.2)	18.1 (0.0)	6.1 (0.0)	6.2	797.9 (0.4)	7.4	108 (0.7)		101.7 (2)
Cipro	Pass	Pass	Pass	Pass							100.8 (0.8)		102.6 (1.7)
Norfloxacin 400 mg											Nor		Nor
NTJ043	Pass	Pass	Pass	Pass	123.1 (11.7)		12.1 (0.0)	19.7	577.1 (1.4)	3.7	101.8 (0.9)		19.9 (6.7)
NTJ044	Pass	Pass	Pass	Pass	121.2 (9.9)		12.7 (0.0)	0.7	651.5 (1.1)	2.9	101.8 (0.6)		88.8 (2.8)
NTJ045	Pass	Pass	Pass	Pass	192.7 (15.6)		18.0 (0.1)	6.8	684.3 (1.1)	2.5	100.4 (0.3)		106.6 (1.2)
NTJ046	Pass	Pass	Pass	Pass	155 (27.6)		15.6 (0.1)	1.4	575.9 (1.0)	2	99.8 (0.5)		94.8 (4.8)
NTW047	Pass	Pass	Pass	Pass	134.8 (12.0)		17.1 (0.0)	5	578.5 (0.4)	1	100.9 (0.8)		81.5 (4)
NTW048	Pass	Pass	Pass	Pass	132.3 (12.7)		15.5 (0.0)	1.2	579.5 (0.7)	2.7	97.5 (0.4)		105.3 (0.7)
NTD049	Pass	Pass	Pass	Pass	165.0 (21.3)		18 (0.1)	4.3	680.0 (0.8)	2.1	98.4 (1.1)		110.3 (1.1)
NTD050	Pass	Pass	Pass	Pass	213.1 (12.3)		12.6 (0.1)	2	650.0 (1.1)	4.1	96.9 (1.2)		97.6 (1.5)
NTD051	Pass	Pass	Pass	Pass	285.6 (4.2)		15.6 (0.0)	4.6	590.0 (1.5)	5.9	96.2 (0.5)		95.6 (5.1)
NTD052	Pass	Pass	Pass	Pass	170.2 (17.8)		12.0 (0.1)	3.4	560.0 (0.8)	2	100.8 (0.5)		104.8 (0.8)
NTD053	Pass	Pass	Pass	Pass	162.5 (9.9)		14.2 (0.0)	1.7	540.0 (0.5)	2.1	97.6 (0.4)		97 (3.1)
NTD054	Pass	Pass	Pass	Pass	161.6 (6.5)		14.2 (0.0)	2.5	540.0 (0.4)	1	99.1 (1.0)		98.5 (2.3)

Amox = amoxicillin; Augm = augmentin; AV = acceptance value; Cipro = ciprofloxacin; Clav = clavulanic; DT = disintegration time; GPHF = Global Pharma Health Fund; I = identification; Nor = Norfloxacin; PDR = percent drug release (dissolution); Q = semiquantitative GPHF thin layer chromatography spots observation; VI = visual inspection.

All tablet formulation samples tested for hardness, thickness, and diameter met the minimum requirement of hardness (40 N)²² and the U.S. FDA recommendations for thickness and diameter.²⁷ The mean hardness, thickness, and diameter results of 10 tablets of each sample with their respective standard deviations are presented in Table 2. All the tested samples were observed to have a mean thickness and diameter within 5% of the standard deviation, showing satisfactory manufacturing control over their thickness.

The USP specifies that tablets and capsules should disintegrate after a specific period of time—15 minutes for uncoated tablets and 30 minutes for film-coated tablets and hard gelatin capsules.²² All the amoxicillin capsules were hard gelatin formulations, and the remaining samples were film-coated tablets. All samples complied with the disintegration time requirement. The results of the disintegration test for amoxicillin are provided in Table 1, and those of amoxicillin/clavulanate, ciprofloxacin, and norfloxacin are provided in Table 2.

In this study, all 54 samples and two comparator products were assessed for the percentage content of the API(s). All the samples and the comparator product of ciprofloxacin and norfloxacin met the requirement for percentage content of the API. The five samples of amoxicillin that were suspected in the GPHF-Minilab TLC quantitation also failed the HPLC content analysis, with a maximum content of 65.5% (sample ACJ007) and a minimum content of 58.7% (sample ACD017). Two samples of amoxicillin/clavulanic acid tablets did not comply with one of the active ingredients with respect to the assay. The sample ACTJ024 contained 89. 3% amoxicillin, while the sample ACTW025 that passed the TLC screening test contained only 44.7% clavulanic acid. One of the samples (ACTD026) that was suspect in the TLC analysis was found to contain both active substances within the USP limit.

Only those samples that passed the assay test were subjected to the dissolution analysis. Seventeen samples of amoxicillin capsules, one comparator and five samples of amoxicillin/clavulanic acid tablets, one comparator and 15 samples of ciprofloxacin tablets, and 12 samples of norfloxacin were tested for in vitro dissolution. The calibration curves for the respective reference standards, shown in Supplemental Figure 3A–D, with their respective linear regression equations and coefficients of determination values showed linearity in the concentration ranges analyzed. The linear regression equations generated were used to determine the amounts dissolved at each sampling time point of the respective drugs. The results of the pharmacopoeial dissolution test data are presented in Tables 1 and 2. The results of this study showed that 12.2% of (six out of 49) samples (one amoxicillin capsule, two from each of amoxicillin/clavulanate and ciprofloxacin, and one of norfloxacin) tested did not meet pharmacopoeial specifications of USP and BP (for norfloxacin). The second and third stage of dissolution test was not performed.

The in vitro dissolution profiles of 16 samples of amoxicillin 500 mg capsule and its comparator product (ACD015) in distilled water are shown in Figure 3. All the samples released more than 90% of the drug within 60 minutes, except sample ACJ005, which released less than 70% of the drug throughout the whole test duration. The USP/National Formulary specifies that the amount of amoxicillin released after 60 minutes should not be less than 80% of the stated amount. An in vitro mean drug release at each sampling point for amoxicillin is provided in Supplemental Table 3.

Figure 3.

Dissolution profile of amoxicillin capsules using distilled water as a medium. ACD015 is the comparator product (CP).

A comparison of the in vitro dissolution data of amoxicillin capsules showed that 56.2% (nine of 16) of the samples were not in vitro equivalent to the comparator product and could not be used interchangeably. Only two samples of amoxicillin (ACW010 and ACW012) complied for both similarity and difference factor values, and seven were rapid release (releasing > 85% of the drug within 15 minutes). The data for similarity and difference factors are also presented in Supplemental Table 3.

The dissolution profiles of five amoxicillin/clavulanate tablet samples and one comparator product are shown in Figure 4. Within the first 20 minutes, all the samples and the comparator product released more than 85% of amoxicillin, satisfying the USP specification—that is, amoxicillin and clavulanic acid combination tablets should release NLT 85% (Q) of the labeled amount of amoxicillin and NLT 80% (Q) of the labeled amount clavulanic acid after 30 minutes for single point dissolution test. All the samples, except for ACTJ024 and ACTW025, and the comparator released more than 95% of clavulanic acid during the first 30 minutes. The percentage of clavulanic acid dissolved remained below 80% until the final sampling point for samples ACTJ024 and ACTW025. An in vitro mean drug release at each sampling time for amoxicillin/clavulanate is provided in Supplemental Table 4.

Figure 4.

Dissolution profile of different brands of amoxicillin/clavulanate 625-mg tablets in distilled water.

The in vitro dissolution profile comparison of amoxicillin/clavulanate showed that 60% (three of five) of the samples were not equivalent to the comparator product, and the data are presented in Supplemental Table 4. Only one sample (ACTJ023) of amoxicillin/clavulanate was found to be similar to the comparator product for both APIs in both factors; the other two products were rapidly releasing both APIs.

The dissolution profile of 15 samples and one comparator product of ciprofloxacin tablets in 0.01N HCl medium and the results are presented in Figure 5. According to the USP, at least 80% of the indicated amount of ciprofloxacin should be released within 30 minutes. Three of the samples (CTJ030, CTJ031, and CTD040) were slow release, whereas the majority of the samples (12 of the samples and the comparator product) released more than 85% of the labeled amount of ciprofloxacin within the first 15 minutes. An in vitro mean drug release at each sampling minute for ciprofloxacin is provided in Supplemental Table 5.

Figure 5.

Dissolution profiles of different brands of ciprofloxacin HCl in 0.01N HCl as a medium.

Analysis of the similarity of the dissolution data of ciprofloxacin showed that 20% (three of 15) of ciprofloxacin multisource products are not in vitro equivalent to the comparator product, and the data are shown in Supplemental Table 5. Six of the products met both similarity and difference factors, and the other six were rapid-release products that do not require further profile comparison to show their equivalence.

Dissolution profiles of 12 samples of norfloxacin tablets in pH 4.0 acetate buffer using the BP method are shown in Figure 6. According to the BP, at least 80% of the specified amount of norfloxacin should be released within 30 minutes; as shown in the figure, norfloxacin samples followed varying drug release patterns at different time points. Sample NTJ043 showed the slowest drug release trend during the entire sampling period, and sample NTW047 showed the second slowest drug release profile; both are manufactured locally by the same manufacturer with different batch numbers. The slow release of sample NTJ043 is in line with its longest disintegration time (19.73 minutes, as shown in Table 2). An in vitro mean drug release at each sampling interval for norfloxacin is provided in Supplemental Table 6.

Figure 6.

Dissolution profile of different brands of norfloxacin in pH 4.0 acetate buffer. NTD052 is the comparator product (CP).

Dissolution profiles of norfloxacin samples was also assessed for similarity with the comparator product using similarity (f₂) and difference (f₁) factors, and the findings are shown in Supplemental Table 6. Six of 11 (54.5%) norfloxacin samples were found to be pharmaceutically nonequivalent with respective comparator products. Only sample NTW048 showed values of f₂ and f₁ within the acceptable range. Seven of the samples (NTJ045, NTW048, NTD049, NTD050, NTD052, NTD053, and NTD054) released more than 85% of norfloxacin during the first 15 minutes. The in vitro dissolution data of all the products tested in this study were fitted into various mathematical models in DDSolver software. The dissolution data modeling showed that the Weibull kinetic model was found to be the best fit for the dissolution data of the reference and all samples of amoxicillin. The Makoid–Banakar model also properly describes the release kinetics of all samples except for sample ACJ005. Samples ACJ006, ACJ009, ACW010, ACW012, and ACD018 can also be explained by first-order kinetics, whereas the dissolution data of sample ACW010 can also be explained by the Hixson–Crowell model. The dissolution data of the comparator product (sample ACD015) was also properly described by the Higuchi model. However, among the models used, the Weibull model best fits the dissolution data of all samples and the comparator product, so it can be concluded that all the studied samples of amoxicillin capsules follow similar mechanisms of drug release. The model fitting of the dissolution data of amoxicillin/clavulanic acid tablets also showed that the first-order kinetic model can properly fit the dissolution data of amoxicillin for samples ACTW025 and ACTD026, and the dissolution data of clavulanic acid for sample ACTD026. The Weibull kinetic model, however, was found to be the best fit for the dissolution data of the reference and all the samples of amoxicillin/clavulanic acid, indicating that the products follow similar drug release mechanisms.

Similarly, the dissolution data of ciprofloxacin hydrochloride tablets were able to be properly described by only two of the kinetic equations according to the model goodness of fit. The Weibull and Makoid–Banakar models were found to best fit the release data of the drug products. However, the Weibull kinetic model was found to be the best fit for the dissolution data of the reference and all samples except samples CTJ029 and CTJ032. Among the models fitted to the dissolution data of norfloxacin, the Makoid–Banakar model and the Weibull model can properly explain the release kinetics of all samples of norfloxacin. In addition to these models, sample NTJ043 can also be described by zero-order kinetics. Overall, the Weibull model best fits the dissolution data of all samples of norfloxacin. Therefore, it can be deduced that all the samples of norfloxacin have similar drug release mechanisms.

The parameters used for the selection of the best kinetic model for the dissolution data of amoxicillin capsules, amoxicillin/clavulanate tablets, ciprofloxacin, and norfloxacin are presented in Supplemental Tables 7 through 10, respectively. The predicted and observed dissolution profiles of reference and test products are presented in Supplemental Figures 4 through 7.

The prevalence of substandard products.

The prevalence of substandard products as analyzed by different factors such as type of facility of sample collection, substandard products by drug product type, stated country of origin, and collection sites is also shown in Figure 7. Of the total 12 substandard products, 7 (58%) were amoxicillin, 2 (17%) were amoxicillin/clavulanate, 2 (17%) were ciprofloxacin, and 1 (8%) was norfloxacin, which is consistent with the finding of the systematic review on SF antimicrobial agents that reported the highest risk of SF products for beta-lactam antibiotics, followed by quinolone antimicrobial agents.¹¹

Figure 7.

Proportion of substandard products.

Three of 11 (20.4%) samples collected from governmental facilities were substandard, failing to comply with the specification for dissolution, and from a total of 43 samples collected from private facilities, nine (20.9%) were substandard. With respect to the location of sample collection, the observed failure rate was 4.3%, 33.3%, and 40% in Dire Dawa, Jijiga, and Togo-Wuchale, respectively.

In addition, the proportion of failed samples (12) per facility type was 8%, 23%, 31%, and 38% for rural drug vendors, hospital pharmacies, community pharmacies, and drug shops, respectively. This finding agrees with the idea that the drug supply chain is most vulnerable toward the point of use, and SF medical products may enter it more easily, as evidenced by comparable findings in Latin America by Rojas-Cortés and Robin,³¹ who reported detection rates of 46.0%, 21.8%, and 10% SF medicines in drug shops, pharmacies, and hospitals, respectively. This level of drug supply chain is not only vulnerable to SF medicines; the degree of regulatory control also decreases as a drug progresses through the supply chain, where the number of stakeholders and transactions increases.⁸ According to the stated country of origin, 59%, 17%, 8%, 8%, and 8% of the substandard products evaluated in this study were stated to be from China, India, Kenya, Turkey, and locally manufactured, respectively. Overall, 67% of the substandard products are stated to be from Asia, and 16% are from Africa, which agrees with the systematic review report of Kelesidis and Falagas, which states the majority of SF antimicrobials originate from Asia and Africa.¹¹

DISCUSSION

The results of this study showed that out of a total of 54 samples collected from public and private hospitals, pharmacies, drug stores, and rural drug vendors within Dire Dawa, Jijiga, and Togo-Wuchale, 22.2% (12/54) of the drugs failed to comply with one or more of the compendial quality specifications. All the samples contained the stated APIs, and no falsified product in terms of API absenteeism was detected in this study. Poor-quality medications have a negative impact on patients, including treatment failure and toxicity. Therefore, the quality of medicines should be consistently assured so that they can be used safely. The finding of this study is comparable with the 18.7% estimate for the prevalence of SF medications given by Ozawa et al.,⁹ from a meta-analysis of more than 40 medicine-quality studies in Africa. Similar results were also reported in a study on the quality of antimicrobial medicines in Cameroon and the Democratic Republic of Congo, which revealed an overall failure rate of 18.6%.¹²

The visual inspection of the product and its packaging, often promoted as a first step in the detection of SF drugs, provides qualitative data on the quality and identity of the drug. A visual inspection of the samples showed no sign of falsification. The registration status of the products was evaluated during visual inspection, and 14.8% of the samples were not approved for marketing by the EFDA, raising the possibility that they could be SF, as demonstrated by the results of the compendial quality study analysis. With the exception of one sample, none of the unregistered products passed the pharmacopoeial quality tests. Because the quality, efficacy, and safety of unregistered medicines are unknown, their availability on the market is a cause for public health concern. The use of medications that have not been subjected to a regulatory assessment and approval process has been linked to an increased incidence of adverse drug responses or a lack of therapeutic effect.³² Therefore, to safeguard public health, it is important to effectively and efficiently regulate medical products on the market and ensure that they have undergone a proper review and approval process.

GPHF-Minilab screening of all samples of amoxicillin, amoxicillin/clavulanic acid, and ciprofloxacin showed that the products contain the stated APIs. However, 14.3% (six of 42) of the samples failed the screening test for the criteria of semiquantitative/quantitative investigations. This is in agreement with a previous study conducted on amoxicillin, amoxicillin/clavulanate, and ciprofloxacin, which reported that 15 of 451 samples failed the screening test.¹² One of the samples in this study that failed met pharmacopoeial quality specifications, and similar results were also reported previously corroborating the low sensitivity of the GPHF-Minilab in detecting some drugs.¹²

All the samples assessed in this study complied with the corresponding minimum requirements for hardness, thickness, diameter, and disintegration time. However, the majority of the samples had hardness values over 100 N despite meeting pharmacopoeial quality specifications for disintegration time. The longest disintegration times observed in two of the samples were replicated in a dissolution test, both of which failed the pharmacopoeial limit for dissolution. A similar study conducted in Ethiopia reported that the majority of ciprofloxacin tablets had a mean hardness value above 480 N but acceptable disintegration times.³³ A related result was also reported for amoxicillin/clavulanate tablets with a hardness as high as 330 N but with acceptable disintegration time.³⁴

Percentage of medications failing to meet pharmacopoeial requirements for percentage content of API, dissolution, and uniformity of the dosage units was 13.0% (seven of 56), 12.2% (six of 49), and 11.1% (six of 54), respectively. Similar findings were reported in the WHO Quality of Antimalarial drugs in Subsahran Africa (QAMSA) research, which had examined antimalarial drug quality in six African nations. It found failure rates in assay, dissolution, and uniformity of dose units of 10.9%, 15.0%, and 6.4% magnitude, respectively.³⁵ A drug product that contains the required quantity of APIs but does not dissolve within the appropriate time will not be absorbed through the GIT lumen. Hence, it will have reduced therapeutic efficacy. Moreover, microorganisms exposed to such a subtherapeutic level of antimicrobial drug may also develop resistance.⁷

Analysis of the in vitro equivalence of the multisource products with comparator products using the similarity and difference factors showed that 56.2% (nine of 16) of the samples of amoxicillin, 60% (three of five) of the samples of amoxicillin/clavulanate, 20% (three of 15) of the samples of ciprofloxacin, and 54.5% (six of 11) of the samples of norfloxacin were pharmaceutically nonequivalent with their respective comparator products. These in vitro nonequivalent products can have varying therapeutic efficacy and cannot be used interchangeably. Comparable findings to the current study were also reported in a comparative dissolution analysis of different brands of amoxicillin in Ethiopia, which reported that 62.5% of the products were nonequivalent to the comparator product.³⁶ A similar study in Ethiopia reported that 80% of different brands of amoxicillin were not equivalent to the comparator product.³⁷ A related study in Nigeria that assessed five amoxicillin/clavulanate products reported that 40% of the products are nonequivalent to the comparator product.³⁸ Similarly, data from an in vitro comparative quality study of norfloxacin in Ethiopia, which reported that 85% of a total of eight brands assessed were not equivalent to the comparator product, are in agreement with the current study.³⁹ In contrast, a study conducted in Ethiopia in 2018 using similar experimental conditions as the current study reported 100% equivalence of generic ciprofloxacin tablets with the innovator brand.⁴⁰ Nevertheless, these findings in general indicate that there is a need for regulatory bodies to monitor and oversee the marketing of pharmaceuticals continuously to ensure bioequivalence, for which there is evidence of nonbioequivalence from various firms, leading to efficacy concerns. Bioequivalence studies that involve in vivo methods can be waived based on the solubility and gastrointestinal permeability of drug substance and can be strategically deployed to save time and resources during generic drug development. In this case, performing in vitro dissolution studies and comparing the similarity and difference factors will suffice. Amoxicillin is in Biopharmaceutical Classification System (BCS) class I; amoxicillin and clavulanic acid are in BCS class II, and norfloxacin and ciprofloxacin are in BCS class IV.⁴¹

Different mathematical models can be used to describe the dissolution data of different products. However, the Weibull model best fits the dissolution data of all samples and the comparator products, indicating that the majority of the products follow similar drug release mechanisms. The Weibull model is not mechanistically realistic, but it has been validated with in vivo functional performance of immediate-release solid oral dosage formulations.⁴² The model was also reported to have been successfully applied to the in vitro dissolution data of norfloxacin immediate release tablets in southwest Ethiopia.³⁹

Limitations.

The limitations of this study are the small number of samples in terms of other antibiotics, the geographic area coverage, and the convenience sampling technique used in this study. Therefore, prevalence estimates generated may not generalize to other medicines and areas. However, the results provide an initial indication of the problem. These data can be used as a baseline for subsequent surveys. The in vitro comparative analysis also did not cover the recommended three pH media, and in vivo studies were not performed. Nevertheless, the data in general indicate a need for regulatory bodies to monitor and oversee the marketing of pharmaceuticals continuously to ensure bioequivalence, safety, and efficacy.

CONCLUSION

Overall, 22.2% (12 of 54) of the products analyzed did not meet pharmacopoeial specifications, indicating that these products are of substandard quality. Substandard and unregistered products were found in this study. Most samples were found to be pharmaceutically nonequivalent with their respective comparator products. Therefore, health professionals should avoid assuming that formulations supplied across national borders are therapeutically similar, even if they are labeled with the same pharmaceutical active ingredient and strength. The assessed drugs are antibacterials, which are prescribed for lifesaving treatment; thus, poor-quality medicines and nonequivalent formulations have a great impact on public health and therapeutic outcome. Such drugs may also be a contributing factor to antimicrobial resistance. Hence the regulatory authority and other stakeholders should devise mechanisms to enforce medicine regulation in porous border areas of Ethiopia.

Note: Supplemental material appears at www.ajtmh.org.