Objective. To estimate the proportion of pharmacies meeting licensing requirements and to identify factors associated with these pharmacies in urban Rawalpindi, Pakistan.
Design. Cross-sectional questionnaire survey conducted during July–September 2001, of 311 pharmacies selected from a drug company list of 506.
Setting. Free-standing licensed and unlicensed pharmacies in urban Rawalpindi.
Study participants. A pharmacist or (if unavailable) the most experienced drug seller.
Results. The proportion of pharmacies meeting licensing requirements was 19.3% [95% C.I (confidence interval): 15.1, 24.2], with few qualified persons (22%). Only 10% had a temperature-monitoring device and 4% an alternative power supply for refrigerators (present in 76% of pharmacies). Associated with pharmacies meeting licensing requirements was the knowledge of not giving co-trimoxazole, a prescription drug, without prescription [OR (odds ratio) = 2.0; 95% CI: 1.1, 3.6], knowledge of the temperature range for vaccines (OR = 2.6; 95% CI: 1.4, 4.8), availability of vaccines (OR = 2.8; 95% CI: 2.8, 18.4), and alternative power supply for the refrigerator (OR = 6.0; 95% CI: 1.5, 23.7). The practice of selling drugs without prescription was not found to have a significant association (OR = 1.1; 95% CI: 0.5, 2.3); however, it did show a trend indicating discrepancy between knowledge and practice.
Conclusions. Most drug sellers had fragmentary knowledge regarding drug dispensing and storage, and improper dispensing practices. There is a need to enforce existing legislation with training programmes directed towards drug sellers and to involve the pharmaceutical industry, which plays an important role in influencing pharmacy knowledge and practices.
Drugs and vaccines are important tools in combating diseases worldwide, but may also cause adverse events, with severity depending upon patient- and product-specific factors. This reality is even more important in developing countries where most medications can be purchased over the counter, including those with a high incidence of side effects or adverse reactions [1,2]. When coupled with a greater propensity towards self-medication, this may have serious consequences for public health in developing countries. National data on self-medication are not available in Pakistan, but local studies reveal that this may vary between 6.3% and 51.3% depending on the setting [3,4]. However, self-medication with appropriate guidance can also be beneficial, keeping in mind the limited resources and health facilities available in developing countries. In this context the role of pharmacies has broadened, incorporating not only dispensing, but also health education, and sometimes even diagnosis.
In the developing world, drug sellers operating in the informal sector are often the first source of health care, especially in settings where self-medication is the norm . Reasons cited by patients include expediency, convenience, efficacy of medicines, dependability of supply, and cost reduction . This has important implications for the role of drug sellers, more so for those with no qualifications in pharmacy. Unfortunately many drug sellers tend to recommend medicines with lucrative profit margins without sufficient concern for the appropriateness (indications, efficacy, safety) of the drug .
A preliminary study conducted in Pakistan reported that many drug sellers have minimal formal education and little or no professional training; of those with training, most were absent from pharmacies , a practice also observed in other developing countries [2,5,8–10]. Given the limited knowledge of sellers regarding indications, contraindications, and side effects, their dispensing practices may have undesired effects.
Although there is a network of health services in Pakistan’s public sector, and a plethora of private sector initiatives, 45% of the population still lacks access to health services . The physician:population ratio is 57 per 100 000 people as compared with 85 in Iran or 311 in developed countries like Sweden . To meet health needs and to reduce personal costs, people rely on alternative health care systems such as chemists, traditional medicine practitioners, faith healers, and homeopaths. Large gaps in the formal health sector encourage self-medication; hence the knowledge and practices of drug sellers becomes critically important in Pakistan.
There are an estimated 45 000–50 000 retail and wholesale drug outlets in Pakistan (population 145 million). However, only about 800 individuals per annum graduate as Bachelors of Pharmacy, which translates to very few qualified personnel available to staff these outlets. A majority of drug stores, therefore, have persons with little or no professional training.
The Government of Pakistan (GOP) has established preconditions for issuing a pharmacy licence. The drug laws of Pakistan  require that:
pharmacy premises should have proper and adequate facilities for drug storage including refrigeration,
premises be clean and in a hygienic and tidy condition, and
drug sales should be under continuous personal supervision of a pharmacist (Pharmacy Act, 1967).
Given that Pakistan has insufficient data on the operation of pharmacies, this study was undertaken to estimate the proportion of Rawalpindi pharmacies meeting GOP licensing requirements, and to identify factors associated with meeting these requirements
Rawalpindi (population 3.4 million) is situated 280 kilometres from Lahore and 173 kilometres from Peshawar. It shares an international airport, railway and bus stations with Islamabad, Murree, and Peshawar. There are 16 hospitals and 200 basic health facilities.
As no GOP list of pharmacies was available, a cross-sectional survey was conducted in Rawalpindi from July to September 2001, using a list provided by a multinational pharmaceutical company. The list consisted of all free-standing pharmacies visited by their representatives (effective 16 May 2001) including licensed and unlicensed operations. While we cannot be sure that the list is comprehensive, we believe that it includes the majority of such facilities in Rawalpindi.
The sampling frame included any outlets in urban Rawalpindi selling allopathic medicines and homeopathic or herbal medicines if sold alongside allopathic medicines. Excluded were pharmacies located on hospital premises or where a doctor was practising (i.e. private dispensaries), and those selling only homeopathic or herbal medicines. Any shop meeting this definition constituted the sampling unit, the sampling element being the drug seller. A simple random sample of 310 pharmacies was obtained using a random number table. From each outlet, a pharmacist with a Bachelor’s degree or Diploma in Pharmacy was identified . If a pharmacist was unavailable, the most experienced drug seller was interviewed, a single respondent thereby being selected from each outlet.
For the purpose of sample size determination, due to the paucity of studies on this topic in Pakistan, the proportion of pharmacies meeting licensing requirements was assumed to be 50% and, with a population size of 506 (after applying exclusion criteria), the sample size calculated was 224. For associated factors, assuming the prevalence of an unqualified proprietor in pharmacies meeting licensing requirements to be 29% , a sample size of 310 was determined suitable to meet both study objectives. Ethical approval was obtained from the Aga Khan University’s Ethical Review Committee.
Training in the use of structured questionnaires was conducted over a 7 day period in early July 2001. Pre-testing was performed in the second week on 31 questionnaires, or 10% of the sample size. Formal data collection ran from 15 July to 30 September involving two data collectors and the principal investigator. Permission for the study was obtained from the District Health Officer. Study subjects were briefed and informed verbal consent obtained. To ensure quality control, completed questionnaires were screened daily by the principal investigator. After data collection, data were edited to ensure quality for data entry.
The questionnaire collected demographic data on the drug seller’s age, sex, years of formal schooling, professional qualification, and years working in a pharmacy. It included fields related to professional knowledge and practice: sources of current drug information, whether certain drugs could be sold without prescription, temperature range for vaccines, commonly used terminology, and dispensing practices. Information collected on pharmacy characteristics included: location, availability of vaccines, whether selling drugs exclusively or general goods also, and presence of an alternative power supply for the refrigerator.
The rationale for this selection of variables was to represent the person, the product, and the environment, all three of which were considered equally important to drug and vaccine efficacy and safety. The condition of the pharmacy floor (used as a proxy for cleanliness) was assessed by its smoothness and whether it was washable. A pharmacy obtaining the maximum score of 3 was labelled as meeting licensing requirements, while a pharmacy obtaining a score of either 0, 1, or 2 was labelled as not meeting requirements.
All data were double-entered in EPI INFO version 6.04 and analysis performed in SPSS version 10.0. Descriptive statistics were generated to ascertain patterns of distribution. For logistic regression analysis, the outcome variable (‘pharmacies meeting licensing requirements’) was devised by summing three variables: presence of a pharmacist, presence of a refrigerator, and condition of the floor (adequacy of person, product, and environment) and are based on the requirements under the drug laws of Pakistan .
In univariate analysis, the unadjusted association of factors with the outcome variable was evaluated; odds ratios (ORs) and 95% confidence intervals (CIs) were calculated. Appropriate scales for continuous variables (age, work experience, years of formal schooling) were assessed through quartile analysis.
Multivariable analysis served to assess the association of factors with pharmacies meeting requirements while adjusting for confounding effects of other variables. Adjusted ORs (adjOR) and their 95% CIs were used to interpret the model.
A total of 311 pharmacies was enrolled; interview response rate was 100%. Out of 311, only 60 (19.3%, 95% CI: 15.1, 24.2) pharmacies situated in urban Rawalpindi met licensing requirements.
The mean age of drug sellers was 39 years (SD 8.04), nearly all (99%) males (Table 1). Only 22% were qualified pharmacists; mean years of formal schooling was 12.6 (SD 1.6). The majority of drug sellers (71%) obtained their current information about drugs from industry representatives. The mean years of work experience of drug sellers was 12.8 (SD 7.0).
Distribution of basic characteristics of the drug sellers in pharmacies situated in urban Rawalpindi (n = 311)
Age (years) mean (SD)
Bachelor of Pharmacy
Diploma in Pharmacy
No pharmaceutical education
Years of formal schooling, mean (SD)
Source of current drug knowledge
On the job experience
Neighbouring chemist shop
On the job experience and medical representative
Books and medical representative
Medical representative and doctor
On the job experience and medical representative and doctor
Journal articles and medical representative and doctor
Work experience of the drug seller (years of work in a pharmacy), mean (SD)
SD, standard deviation.
About 60% of pharmacies sold only drugs, while the remaining 40% sold other items as well (Table 2). The majority (76%) had a refrigerator present; however, only 10% of these had a temperature-monitoring device, and only 4% had an alternative power supply. Another legal requirement is that the licence should be displayed openly, but this was visible in only 6% of pharmacies. In nearly all (97%), the floor was found to be smooth and washable. The majority (70%) were located away from a hospital. Vaccines were available in 60%.
Among the most commonly bought antibiotics, quinolones and cephalosporins comprised the major proportion, followed by penicillins, co-trimoxazole, and tetracyclines. A large proportion of drug sellers (64%) thought (incorrectly) that co-trimoxazole could be dispensed without prescription (Table 3). Only half (53.7%) had correct knowledge about the holding temperature for vaccines. Nearly all could interpret some prescription terminology: OD, BD, and TID; however, knowledge regarding QID, HS, and SOS was unsatisfactory. When sellers were asked whether they sold prescription drugs without prescription, around 16% admitted to this.
OD, once a day; BD, twice a day; TID, three times a day; QID, four times a day; HS, before going to bed; SOS, when required.
Univariate analysis was performed to assess the association of independent variables with pharmacies meeting licensing requirements. Pharmacies meeting requirements were more likely to be situated near a hospital, to have vaccines and an alternative power supply for the refrigerator (Table 4). These were also more likely to have drug sellers with: work experience of 16 years or more, a variety of sources for acquiring current drug knowledge other than representatives, knowledge that co-trimoxazole cannot be given without prescription, knowledge of the temperature range for vaccines, and knowledge of prescription terminology, than pharmacies not meeting requirements. However, the practice of giving drugs without prescription was not found to have a significant association.
The final logistic regression model included: availability of vaccines, alternative power supply for the refrigerator, drug seller’s work experience, knowledge of vaccine storage temperatures, knowledge that co-trimoxazole requires prescription, and the practice of giving drugs without prescription (Table 4), and was selected on the basis of statistical significance and biological plausibility.
After adjusting for the effect of other variables, pharmacies meeting requirements were more likely to have vaccines available and an alternative power supply for the refrigerator. Drug sellers with: work experience 9–11 years, 12–15 years, and 16 or more years, knowledge of temperature range for vaccines, knowledge that co-trimoxazole should not be given without prescription were associated with meeting requirements after adjusting for all other variables; interestingly, they were also more likely to give drugs without prescription.
Only about one-fifth of pharmacies sampled met licensing requirements. Most had unqualified personnel selling drugs and providing advice and treatment on common health problems, while drugs were commonly sold irrespective of whether the client had a prescription, as documented in other developing country studies [7,9,10,14,15]. This is alarming yet not surprising, as drug sellers play a major role in prescribing in Pakistan, where nearly half the population lacks access to qualified medical practitioners. Furthermore, most drug sellers obtained their drug information from industry representatives, also as reported from other developing country studies [8,16].However, the promotional strategy of representatives, extolling a drug’s benefits and minimizing side effects, is largely dictated by financial incentives to promote sales [6,17–19]. Such marketing practices are well documented in both developing and developed countries; in the USA drug company enticements were linked to increased prescribing of the promoted drugs , physicians relying on representatives as their source of information prescribe more drugs and more often prescribe inappropriately . This commercial relationship between physicians, drug sellers and representatives plays a major role in influencing pharmaceutical practice .
Nearly half of pharmacies were found selling general goods like foodstuff, cosmetics, and miscellaneous household items along with drugs. As pharmacies can be kept open for 24 hours, it is likely that in many of these instances, drugs are kept primarily as an adjunct to items of general use even without a pharmacist present, reflecting predominantly commercial rather than health care motivations.
Our study also examined storage practices, revealing a paucity of temperature monitoring devices and alternative power supplies for refrigerators, and raising doubt regarding the efficacy of products requiring refrigeration, as also reported from India . More than half of pharmacies were keeping vaccines irrespective of appropriate storage, consistent with an earlier Karachi study , with serious implications for vaccine efficacy and the ultimate effectiveness of vaccination initiatives, at least to the extent to which population coverage is dependent on private sector suppliers.
The assessment of baseline knowledge and practices of drug sellers revealed that over half were unaware that co-trimoxazole is to be sold on prescription. This drug is often used as an indicator when enquiring about knowledge of antimicrobials . However (although once responsible for a high sales volume in Pakistan), co-trimoxazole’s proportional sales were low compared with quinolones and cephalosporins. A possible explanation could be that sale of a drug in Pakistan is largely dictated by its profit margin. For example, quinolones have over 50 competitors in the market with prices ranging from Rs.10 (1 US$ = 60 Pak. Rs., June 2002) to Rs.31 per tablet/capsule while cephalosporins have prices ranging from Rs.10 to Rs.20, and it is in the economic interest of drug sellers to sell those brand names that have a larger profit margin. The underlying public health issue however, is that cavalier use of antimicrobials is a major factor in the emergence of resistant organisms, thereby undermining the efficacy of the drugs themselves.
Over half of the respondents knew the correct temperature range for vaccines and also knew prescription terminology. Only one-fifth admitted to giving drugs without prescription, which may be an underestimate, given that this contravenes correct practices.
Drug sellers with knowledge of the correct temperature range for vaccines and knowledge that co-trimoxazole should not be given without prescription were significantly associated with pharmacies meeting requirements. However, these also were more likely to give drugs without prescription; although not statistically significant, this reveals a discrepancy between knowledge and practice. Similar behaviour is also reported from Ghana , while studies using the Simulated Client Method (SCM)  in Vietnam and Nigeria have also identified this discrepancy [25,26], most likely reflecting an interplay of factors influencing behaviour: refusal of a client’s request may lead to loss of business, while competing drug sellers also resort to such practices to increase profits.
Pharmacies are often the first point of contact for patients seeking health care as they are usually more accessible and less socially distant than other providers, including medical doctors. Attempts to extend the role of pharmacists from dispensers to practitioners are reported from South Africa, and also the UK where they are used in detecting hypertension [27,28]. A study in Ghana identified pharmacies as an accessible avenue of care in managing sexually transmitted infections . However, if drugs or vaccines are compromised by quality issues such as we have documented, households may end up with products or advice that are worthless and even hazardous, a false economy indeed.
Our study had limitations, but allows critical issues to be raised relating to the public health effectiveness and safety of drug and vaccine use in Pakistan. Due to the sensitivity of some questions, not all answers given by drug sellers may be consistent with actual practices. It is for this reason that the market presence of counterfeit drugs could not be addressed using our survey method, as responses to such a question would certainly be biased. Qualitative investigations (such as SCM) are required to explore such issues. We acknowledge that the pharmaceutical company list (used due to the absence of any more formal listing) may not include all pharmacies in urban Rawalpindi and could lead to a bias in sampling, although a large majority of outlets are believed to be represented.
In many areas of public health, education holds more promise for long-term change than does the enforcement of legislation. In this respect, creating avenues by which sellers could enhance and maintain their competence and the respect in which they are held in their communities, may be the best long-term strategy. Such approaches have been advocated and attempted abroad, such as in Ghana and Kenya [29,30], and need to be tried in Pakistan, especially as (for example) the importance of refrigeration with back-up power may simply not be understood by some commercial drug sellers.
In conclusion, involving drug sellers in diagnosis and treatment, or as providers of advice for illnesses is not necessarily an ideal situation. However, in Pakistan, they play a critical role in reaching people in need, and have the potential to provide an accessible, reasonably private, and rapid service. In this sense they are an underutilized community resource that can be enhanced to improve health care.
There are implications for the GOP, with whom we are in ongoing communication regarding the findings of this study. As most pharmacies did not meet requirements, existing legislation should be enforced, and licence renewal linked to an inspection process. Incentives should be introduced to improve service quality. In the longer term, training programmes for drug sellers are needed to enhance knowledge and promote safer practices. More in-depth studies are warranted to further assess the situation of Pakistan’s pharmacies and to improve the quality of their service. Finally, no intervention will be successful without the pharmaceutical industry’s involvement.
The authors thank the Department of Community Health Sciences, the Aga Khan University, for funding this study and are also grateful to the District Health Officer, Rawalpindi for his support and Pharmacia Upjohn for providing the list of drug sellers.