Practices of dispensing doctors - drug use and health economics Trap, B Zimbabwe Essential Drugs Action Programme, Ministry of Health & Child Welfare Problem Statement: Doctors dispense drugs for many reasons, such as insufficient pharmacy coverage, to increased drug accessibility and availability, and for economic reasons. For policies and strategies to be developed, information regarding the rationality of having drugs prescribed and dispensed by one and the same person need to be reviewed and evaluated. Objectives: To review and evaluate the appropriateness and validity of available information pertaining to the practice of dispensing doctors related to the aims of national drug policies in ensuring rational drug use and affordability. Designing and Setting: The review was prepared based on available literature and information obtained at national and international levels. A critical evaluation of the existing literature was carried out and areas for further investigation identified. More than 50 references were reviewed and the various issues were discussed and presented in a coherent way. Outcome Measure(s) and Results: There is little solid information available to develop national policies on the practices of dispensing doctors. Very few valid studies have been undertaken to evaluate the rationality of this practice, and where studies have been done, there is no concurrence through repeated or comparable studies. In addition most of these studies, which examine aspects such as procurement, prescribing and dispensing behaviours, and compliance and costs, have been done in developed countries. The conclusions that have been drawn, in comparison to nondispensing doctors, are that dispensing doctors prescribe more drugs, fewer generic drugs, and have a greater income. Other aspects of dispensing prescribers that have been reported on are the influence of drug company sales representatives, the use of free samples, storage practices and the conflict of interest that receiving payment for prescribed medications poses. The benefits of a second check by a pharmacist / dispenser - such as reducing adverse drug reactions and improving compliance - could not be proven. Ensuring standards in dispensing and control of pharmaceuticals, and rational drug use aspects like adherence to accepted treatment protocols, were not critically evaluated in the studies. Conclusion: Almost all of these studies were undertaken in developed countries, and only the finding of increased drug use has been confirmed in more than one study. In view of the increasing drug costs experienced in most countries, the need for strengthening national policies and rational drug use, and for undertaking cost evaluations in both the public and private sector, is evident. More information on the practice of dispensing doctors is required if optimal policies are to be developed in this area. Studies in this regard should therefore be encouraged. Summary This article reviews literature concerning the practice of dispensing doctors in comparison to non-dispensing doctors and pharmacists. Dispensing is an important step in ensuring drug accessibility, affordability, safety and rational use. To ensure optimal utilization of limited funds, dispensing must be regulated by appropriate policies and laws. To provide politicians with information enabling them to prioritize and legislate, this review was prepared. The various legal aspects regulating dispensing by doctors have been listed. The spectrum of what is legal is very wide. Many laws focus on regulating dispensing when carried out by pharmacists, whereas dispensing by doctors in several countries is not clearly defined. South Africa, Zimbabwe and some states in the USA have recently started giving priority to enforce and regulate dispensing by doctors. Prior to the 11th century all drug dispensing was carried out by doctors. Thereafter, pharmacists became recognized and started taking over dispensing. This development continued in most countries where doctors were paid a consultation fee. In other countries where a doctor's livelihood depends on the sale of drugs, dispensing by doctors has carried on, i.e. in India, Japan and Pakistan, where about 80% of doctors dispense. In other countries, dispensing doctors only constitute a minority of 5-10 %. However, the number of dispensing doctors is increasing. Actual studies on dispensing doctors are very limited and much of the available information is related to emotions, belief and personal opinions, rather than sound scientific evaluations. The following findings have been based on studies carried out mainly in UK, USA and from an inspection in South Africa. In comparison to doctors who do not dispense, those who do were found to prescribe more drugs annually per patient, prescribe fewer generic drugs, issue more prescriptions (but in lower quantities), have older patients, and have less years in practice. The average number of visits per patient per year was found higher for dispensing doctors' patients. In an inspection of dispensing doctors' practices in South Africa, 46% were found to have inappropriate storage facilities for drugs, 52% were staffed by untrained dispensers, 61% did not provide suitable containers and 27% were unsatisfactorily labelled. The benefits of a second check by a pharmacist, in regard to minimising adverse drug reactions or compliance, could not be proven in a study on elderly admissions to hospital. Only few of the findings have reoccurred in more than one study, and several of them are based on very small sample sizes. There is a need to obtain more information on this practice and to repeat previous findings in other countries, including developing countries, in order to ensure the achievement of an optimal health care system catering to the majority of the population. Introduction Doctors dispense drugs for many reasons, such as insufficient pharmacy coverage, to increase drug accessibility and availability, and for economic reasons. For politicians to make policies and strategies to regulate the practice of dispensing by doctors, information regarding the rationality of having drugs prescribed and dispensed by the same person needs to be reviewed and evaluated (33). A definition of a dispensing doctor was offered by SA Axon at the FIP conference in 1993: a dispensing doctor is a medical practitioner who is permitted to provide a dispensing service to some or all of his patients either as an exception to the general legislation governing the provision of pharmaceutical services or as part of the overall provision of medical services (37). The definition relates to the legal restriction of the practice. Depending on the national laws, it allows doctors, in general, to dispense, or it only permits dispensing injections, free samples, in emergencies or in remote rural areas where pharmacy services are unavailable to ensure accessibility (11, 73, 55). The evaluation of dispensing by doctors is far more complex than a simple spirited competition between two or more dispensing alternatives. Drugs cannot be viewed as ordinary commodities of commerce as they are not like most other commodities (72). Drugs are not chosen directly by the buyer (the patient), the buyer is not always the responsible payer, the buyer often has no background or the necessary information to evaluate or make a choice which can have hazardous or even deadly consequences for the patient. Normally market economics have to be modified when applied to drugs as assumptions are not easily fulfilled (77). This has been recognized in regard to pricing, demand and sale of drugs where special models have been developed (77, DK articles). Studies in the UK found that reducing the price of drugs did not markedly affect their use (71). Policy makers must find the best way to achieve the aims of their national health or drug policies and decide on necessary regulation. Part of this is to pass laws on dispensing by doctors. The main objective of the Zimbabwe National Drug Policy is to ensure the availability and accessibility of safe, efficacious, cost-effective, and affordable pharmaceutical products of good quality to the entire population of the country, and furthermore to promote the rational use of drugs through improved information, prescription and compliance (75). Legal aspects Back in 1271 in France, doctors and pharmacists were reported to have feuded over dispensing. This discussion has continued to this day in many countries, developed as well as developing (11, 54, 46, 14, 77). In 1240 the German Emperor Frederick II initiated the separation of the occupations performed by doctors and pharmacists. Doctors became the diagnostic and prescribing experts and pharmacists experts in drug dispensing and drug management. This development became possible due to urbanization whereby markets had become large enough to allow for specialization (11, 38, 39, 59). Doctors could generate an income from diagnosing and prescribing. Likewise, pharmacists were not prescribing but could sustain a living through the sale and dispensing of drugs. The fact that doctors could generate an income from diagnosing and prescribing, and pharmacists from dispensing alone, was fundamental to the separation of the two practices. These principles were accepted by Parliament in the UK in 1912 and by the national health insurance in 1964 (13). Paying for a consultation is, however, not accepted in all cultures. In India and Pakistan patients do not generally accept paying a consultation fee, and the common way doctors generate an income is by dispensing (selling) drugs (13, 62, 63). This is also the practice in Japan, Mexico and South Africa. In 1991, over 80 % of all doctors in these countries dispensed medicines (39b,59, 37). In Japan the law allows dispensing by doctors if "medically necessary". Drug expenses in Japan in 1993 were found to be 30% of the total health care bill, compared to other developed countries which spend about 5% on drugs (37). In 1990 in South Africa, 80 % of drug expenditure was spent in the private sector, although 60-70% of the total volume of pharmaceuticals was consumed in the public sector (74). In countries such as Scandinavia and Germany, the two professions continue to be separate. For example, in Denmark the law does not allow a person to practice medicine and pharmacy at the same time. Doctors dispense injections and drugs only in special circumstances, such as emergencies. Most countries have regulated dispensing by doctors. In the USA 45 of the 50 states have some type of regulation (10). However, the interpretation and implementation of these regulations have met with difficulties due to a lack of resources, ability or authority to inspect dispensing doctors; a lack of clarity on who is responsible for enforcement; and the imprecise wording of the laws (11,10, 65). A US survey in 1989 found that 20 states had strengthened their regulations in the last 2-3 years. However, as mentioned above, the enforcement was hampered because of limited resources, and only a few states conducted inspections of dispensing doctors’ offices (10). A survey undertaken from 1986-1987 found that 20 to 31 US states permitted physician dispensing with no specific or minimal regulations (11). The basic aim in most laws is covered by the following principles: * Rurality: permits physicians to dispense in rural areas where there are no pharmacies. The laws, e.g. in England, Netherlands and Zimbabwe, allow dispensing by doctors if no pharmacy is available with in a mile or 5 km from either the patient's home or, in some laws, from the doctor's practice (37, 73, 56). * Dispensing rights - renewal: once a doctor undertakes dispensing, this right can either be reviewed annually or continues even when a pharmacy is established within the one mile or 5 km limit (73). Compensation for a decrease in income has been much debated in the UK (56). * Emergencies: permit physicians to dispense in emergencies. The laws in 5 US states allow dispensing only in limited situations, such as medical emergencies (10, 59). In Denmark this occurs during an after hours visit to a patient where treatment might be required. * 72 hour supply: in USA the Federal Trade Commission requires that no more than 72 hours' worth of medicine be dispensed by doctors (11, 30). In Germany, since 1993, the law allows doctors to provide treatment at home before and after a stay at hospital in order to shorten an expensive stay at hospital (37, 38). * Not for profit or cost/price regulation: limits the amount physicians may charge for medication they dispense. In Australia, which only has 69 dispensing doctors, passed laws that forbid dispensing doctors from making a profit (37). In 1989 the US congress considered a bill to prohibit practitioners to dispense drugs for profit (10). * Procedural requirements: limit dispensing to physicians' own patients and require dispensing to follow the same regulations mandated for pharmacists, e.g. labeling, record keeping, package requirements and storage. Labeling and record keeping requirements are applied in at least 23 of the 45 states in USA that permit physician dispensing, and in South Africa (10, 11, 59, 65). * Delegation of dispensing: permits dispensing to be undertaken only by trained or employed staff related to the physician's practice. Conflicting laws also exist which require dispensing to be done by the physician personally (37,40). * Patient choice: this principle protects the freedom of choice for patients when deciding whether to purchase prescription drugs from their physician or the pharmacy (8). * Duty at factories: permits physicians to dispense drugs in small amounts to employees during factory visits. This is practiced in Germany (38). * Registration: the laws in Zimbabwe and in 13 states in the US require dispensing doctors to be registered (10, 73). Most laws regulate dispensing to protect public health, safety and welfare. However, interpretation and enforcement often aim to regulate dispensing when carried out by pharmacists, not physicians. However, more countries have started addressing this issue, and drug policies began prioritizing the regulation surrounding dispensing undertaken by practitioners. In 1996 the new National Drug Policy of South Africa stated that dispensing doctors can only be permitted where separate pharmaceutical services are not available. In these cases, the dispensing doctors or dispensing nurses must be registered and annually renew the registration, be trained in dispensing, be inspected, practice good dispensing and show transparency in the pricing structure (74). Development The number of dispensing doctors has recently increased in several countries. In the UK about 12.5% of general practitioners were found to dispense in 1966/67(56). This increased slightly to 14% in 1989-90 and to 15% in 1993(28, 77): dispensing doctors account for about 7% of all primary care prescriptions (28). In the USA, about 40% of all physicians dispensed drugs in 1923; this decreased to 25% in 1947, 8.7-10% in 1967, and about 5%-8.5% in 1990-91 (11, 12, 59). A survey in Louisiana in 1992 based on questionnaires found that only 4 % of physicians dispensed; but the same survey found that 10.5 % stated that they intended to dispense within the next 2 years (30). Other authors also believed that the number of dispensing doctors will continue to increase. The reasons given are to benefit patients and society, and because dispensing offers a possibility to increase or sustain physicians' incomes, which have been declining over the past years. Also, dispensing has been facilitated by the introduction of prepackaged drugs. The decline in physicians' incomes in the USA has been related to increased competition resulting in a fall-off in patient load (59). In South Africa the number of retail pharmacies have remained almost constant from 1985 to 1995, whereas the number of registered dispensing doctors has doubled from about 4000 to almost 9000 in the same 10 years (65). In 1996, new policies were introduced in South Africa aiming at strengthening regulation of dispensing doctors. Similar developments are called for in USA, Zimbabwe and Japan (10, 76, FIP). In countries such as India and Pakistan, dispensing by doctors is the common way for general practitioners to earn a living as this is closely linked to cultural changes which are expected to occur as a result of consumer education (13, 63). Drug dispensing by doctors Doctors dispense for many reasons and the debate has been lively on the pros and cons. Often discussions are related to economics from the pharmacist's or doctor's point of view, giving less priority to the patient and to the aim of the national policies. Pros and cons are difficult to evaluate as very few actual studies have been undertaken regarding dispensing doctors; the discussions have thus been very theoretical and controversial. The heated debate on the practice of dispensing doctors is being waged over a complex variety of ethical, economic, public health and patient care issues (8, 72, 39, 10, 36, 59, 12, 37, 29, 11, 18, 67). To provide information for further studies, the discussion is summarized below as questions related to the outcome of dispensing. Ethics: 1. Is it a conflict of interest for physicians to sell the prescription drugs they prescribe? 2. Is it important that the consumer has the freedom to choose from where to obtain drugs? 3. If doctors are viewed as authorities, do patients feel compelled to buy drugs from them, or will they be able to obtain the drugs from a pharmacy? Would this result in a closed market? 4. Is dispensing by doctors an exploitation of the patient - physician relationship? 5. Will professional ethics and peer pressure protect patients from abuse? Availability: 6. Can availability of drugs be satisfactorily achieved by dispensing doctors? 7. Do dispensing doctors ensure appropriate procurement and distribution from a health economics perspective? Accessibility: 8. Obtaining a prescription and drugs at the same place is convenient and increases accessibility for the patient. Does it outweigh the advantages of dispensing by pharmacists? 9. What is the best system to ensure national accessibility of drugs? Safety: 10. Is dispensing by pharmacists a safety net? Will prescriptions with errors, drug interactions or excessive drug use be prevented or minimized significantly by having the prescription pass through two qualified trained health professionals? 11. How does the use of prepackaged drugs compare to pharmacy dispensing bags with regard to safety, cross contamination, labeling, etc.? 12. How can quality be ensured for drugs dispensed by dispensing doctors versus pharmacist dispensed drugs? 13. Are the laws on dispensing (by doctors or pharmacists) sufficiently structured and implemented to safeguard patients? Affordability: 14. Are drugs distributed through dispensing doctors cheaper or more expensive for the patient and for society? 15. Do dispensing doctors offer the best combination of price and quality of service? 16. Will dispensing by doctors enhance competition and reduce prices? 17. Is a price control structure in place or is affordability based on competition? 18. Must governments ensure the professional status, and provide the necessary background for doctors to maintain an acceptable income? 19. Is there a need for and political interest in having pharmacies in the future? 20. Do dispensing doctors prescribe more brand-name drugs or generic drugs compared to non-dispensing doctors? Rational use: 21. Will dispensing by doctors lead to over prescribing? 22. Will dispensing by doctors narrow the therapeutic options by mainly limiting the drugs prescribed to those in stock? 23. Are dispensing doctors' prescribing habits different from those of non-dispensing doctors, e.g. injections, quantities, frequency, generic prescription and other rational drug use aspects? 24. Do dispensing doctors adhere to good dispensing practices? 25. What is the effect of the physician-patient relationship with regard to rational drug use? 26. How can quality assurance for prescribing best be ensured in the private sector? 27. Can non-health staff be sufficiently trained in dispensing to it take over? 28. Dispensing doctors know what medicine patients are getting when they dispense themselves, and pharmacists cannot change it by generic substitution. Information: 29. Is provided patient information adequate when provided by one health professional instead of two? And is the information unbiased? 30. Is patient knowledge increased or decreased by dispensing doctors? Compliance: 31. Will dispensing by dispensing doctors increase compliance? 32. Can dispensing doctors better ensure prescription refills? Several of these questions can be answered based on scientific evaluations or studies. However only a few of such studies have been undertaken, and many of the above questions still remain unanswered. Factors influencing prescribing by dispensing doctors Good prescribing habits are essential to ensure effective and safe treatment, the shortest duration of illness, less distress and harm to the patient, and lower cost. Bad habits make prescribers vulnerable to influences which can cause irrational prescribing, such as patient pressure, copying bad prescribing habits of colleagues, and pressure from peers and salesmen. The steps in the prescribing process are: defining the patient's problem, specifying the therapeutic objectives, verifying if a drug is to be used, selecting a specific drug (checking effectiveness and safety), choosing dosage and regimen, providing information, writing the prescription or start and monitor treatment. (78, 20). Various factors have been found to influence the prescribing process and determine whether the prescribing is rational. These factors include: Education, diagnostic and therapeutic skills, pharmacological knowledge, advice from colleagues, habits of the "copied prescribers", use of formularies or standard treatment guidelines, years of practice/education, personal clinical experience, patients' demands, prescribers' personalities, marketing activities from drug industry, consultation time, importance of maintaining patient load, quality assurance, control and regulatory measures, information and knowledge on drug costs (71, 62, 69, 20, 78, 29). It is difficult to separate the effect of the individual factors, (71) and only little information is available on the actual effect of the factors. Only very few studies have evaluated dispensing practice carried out by doctors in relation to these factors. Findings Findings from studies comparing variables of dispensing doctors and non-dispensing doctors or pharmacists were given. Not all references had given a detailed method or study description, but only listed results. Where a method was described, it was summarized and commented on when giving the results. Background factors A study examining the differences in the prescribing practices of 59 dispensing and 49 nondispensing doctors from 1989-90 in the UK found a higher proportion of dispensing doctors' patients to be over 65 years old, compared to non-dispensing doctors' patients 0.195 versus 0.176 (28). Distance to pharmacy (km) and rurality have been found to be significantly higher for dispensing doctors than non-dispensing doctors. On average the distance to a pharmacy in the UK was 4.36 km for dispensing doctors compared to 1.0 km for non-dispensing doctors. Rurality (also measured in km), was 2.43 versus 1.25 (28). This is confirmed by a study from California which found a higher percentage of dispensing doctors in rural areas, 12.9% versus 9.9% (11). In Louisiana, where not for profit dispensing is allowed to a maximum of a 72 hours supply, no difference could be found in regard to location between dispensing doctors and non-dispensing doctors (30). In 1992, to evaluate differences between dispensing doctors and non-dispensing doctors, a questionnaire was sent to doctors in Louisiana. Of the 361 who returned the questionnaire, only 13 were dispensing doctors. The survey found no difference between dispensing and non-dispensing doctors with regard to type of practice, organization membership and number of patients seen per week (30). Dispensing doctors in the USA were found to have fewer years in practice compared to nondispensing doctors. 61% of doctors presently dispensing or with intention to dispense had less than 15 years in practice, compared to 39% of non-dispensing doctors (30). The same study also found that dispensing doctors or doctors with intention to dispense were associated with a less positive attitude towards pharmacists. Consultation time was measured in a study on the treatment of childhood diarrhoea in Pakistan, comparing 62 general practitioners who dispensed drugs with 28 non-dispensing pediatricians who were paid a consultation fee. The study included 996 patient encounters and was based on observations of prescribers and patients 3-4 hours daily for 5-6 days. The general practitioners who dispensed drugs spent on average 3 +/- 2 minutes per consultation. 48% of the encounters were less than 2 minutes, compared to non-dispensing pediatricians who spent 9 +/- 4 minutes per patient (63). This difference, however, could also be attributed to the difference in their education. Rational drug use The number of items prescribed annually per patient was found to be 15% higher for dispensing doctors compared to non-dispensing doctors: 9.55 versus 8.32 (28). This might, however, be a result of dispensing doctors prescribing smaller quantities more frequently. Comparing defined daily doses (18) or using the INRUD indicators (79), measuring the number of drugs per encounter is perhaps a more appropriate measurement. A recent study in South Africa in 1996 confirmed that dispensing doctors tend to prescribe more items more frequently. The number of items per script was 2.38 for dispensing doctors compared to 1.67 for non-dispensing doctors (65b). A UK study found the same number of items being prescribed by both dispensing doctors and non-dispensing doctors, namely 2.59 (18). The average number of visits per year was found to be higher for dispensing doctors compared with non-dispensing doctors (3.34 versus 2.48) (65b). The use of injections has been evaluated in Pakistan in relation to childhood diarrhoea, comparing general practitioners who dispense drugs with non-dispensing pediatricians. Doctors' use of injections for dispensing doctors was double that of non-dispensing doctors. 32% and 18% prescribed/dispensed injections in 15% and 8% of their encounters respectively (63). Rational prescribing in regard to dispensing by doctors has been evaluated. The Pakistan study found ORS to be prescribed by 53% of the dispensing doctors compared to 61% of the non-dispensing pediatricians. However, in about 50% of the encounters in both practices, patients were given ORS. Significant differences were found in the use of antibacterials as well as antidiarrhoeals. More dispensing doctors were found to prescribe these types of drugs as compared to the non-dispensing pediatricians: 66% of dispensing doctors prescribed antibacterials compared to 50% of non-dispensing pediatricians; and 60% versus 29% for antidiarrhoeals. Moreover, 89% of the dispensing doctors, in 77% of encounters, dispensed drug formulations of unknown composition and mixtures made in their own drug dispensing areas. History taking was found to be more elaborate for non-dispensing doctors compared to dispensing doctors: 73 versus 63% measured by using four indicator questions. An explanation of these differences could be the dispensing variable, but it should be mentioned that the non-dispensing doctors were specialists and the dispensing doctors were general practitioners (63). Generic prescribing was found to be lower for dispensing doctors in comparison to nondispensing doctors. In the UK, 26.5 % of drugs prescribed by dispensing doctors were generic, compared to 42% for non-dispensing doctors (28). Dispensing doctors in practices that introduced fundholding in 1991 prescribed 27% generics in 1991, increasing to 39% in 1994. This compared to non-dispensing doctors who in 1991 prescribed 44.5 % generics, which increased to 58% in 1994. This increase in generic prescribing in both practices is related to fundholding being an incentive to save on drug expenditures. In 1991, dispensing doctors prescribed cotrimoxazole instead of septrin/bactrim 5% of the time, which increased to 62% in 1994; for non-dispensing doctors the figures are 58% and 97% respectively. The same figures for naproxen versus naprosyn for dispensing doctors were 13.5 % increasing to 66 % in 1994; for non-dispensing doctors, 64.4% increasing to 91% (18, 23). The role of pharmacists to detect errors and interactions is well known (50b). One study showed that 1.6 % of all prescriptions contained errors which were detected by pharmacists. Of these 0.2 % were life threatening (11). Dispensing errors by pharmacy staff have also been surveyed. Based on a 12 day peer-review, errors were detected in 12.4 % of the prescriptions with 1.5% being serious. Drug interactions in other studies were 9.5-22 % (11). It has been argued that pharmacist dispensing only increases the risk of introducing more errors. However, in view of detecting systematic repeated errors, safety increases if two qualified persons are involved in providing drugs to the patient compared to only one (double check). A study compared 905 admissions from dispensing doctors with 3449 admissions from community pharmacists by assessing drug problems in the elderly admitted to hospital. This study however found no significant difference with regard to adverse events (8.4 versus 9.4%), non-compliance (1.8 v 1.3%) and miscompliance for three selected drugs. The authors question the sensitivity of the design in the study (67). Compliance was not evaluated in the specific studies related to dispensing by doctors. However, it was argued that doctors, in comparison to pharmacists, are better at administering eye/ear drops and giving children medicines. The importance of giving an initial dose of treatment and medicine similar to the last received was also highlighted in order to improve compliance (12,15, 38). Furthermore, it is argued that dispensing doctors can ensure better compliance, compared to pharmacists, when refilling prescriptions (11,12,59). Other authors argue that compliance can be better ensured by having the medicines dispensed by pharmacists, where doctors and pharmacist work as a team (11). Dispensing and storage No state in the USA allows a licensed pharmacist to delegate the dispensing function to a nonpharmacist, unless the pharmacist personally supervises the activity. However many laws are ambiguous regarding physicians' authority to delegate dispensing (11). To improve dispensing by dispensing doctors or their assistants, training courses were implemented (40). Based on inspections carried out in South Africa, it was found that in 52% of dispensing doctors' practices, medicines were dispensed by inappropriately trained persons, e.g. receptionists (65). Drug dispensing at pharmacies in Zimbabwe was carried out by pharmacy personnel with only in-service training in dispensing (80). In the same South African evaluation, storage conditions were found to be inappropriate in 39% of dispensing doctors' practices (65). A further survey which included 1103 dispensing doctors' practices in South Africa found unsatisfactory storage conditions for medicine in 46% of practices (64). An expiry date monitoring system was only found in 30% of the practices, and 20% of the medicines were not clearly labeled (64). Adherence to good dispensing practices by dispensing doctors was evaluated. Following an inspection of dispensing doctors' practices in South Africa, 33% were found to count tablets and capsules by hand, 61% did not have suitable containers for dispensing and 27% of practices did not label the dispensed medicine(s) appropriately. The labels did not state expiry date, batch number or name of dispenser (65). Batch numbers for the recall of medicines were not recorded in 61% of practices. In a 1995 study on labeling undertaken by 31 pharmacies in Zimbabwe, labeling was found to be correct in 86% of the cases. The lowest scores were for including the doctor's name on the label (79% correct) and for including the manufacturers name (19% correct) (80). The South African survey also found that patients were given antibiotic powder which they were expected to make up at home (65). The South African study concluded that patient information was unsatisfactory. Patients were not sufficiently informed on how to use medicines, on cheaper generic alternatives or about the unwanted effects of the medicines, what the medicine is expected to do, etc. (65). Information related to the diagnosis was evaluated in the study from Pakistan. Non-dispensing doctors were found to communicate the diagnosis in 37 % versus 11% for dispensing doctors (63). Health economics factors In a study comparing costs per patient in UK using data from the Prescription Pricing Authority, the annual net ingredient cost per patient was found to be 10.3 - 13% higher for dispensing doctors compared to non-dispensing doctors, i.e. GBP 48.47 for non-dispensing doctors compared to GBP 54.78 for dispensing doctors (37, 28). Annual net ingredient cost per item was found to be slightly lower per dispensing doctor patient: 5.74 versus 5.83 (28). To analyze these differences further, 10 indicator drugs were used for comparison purposes, and it was found that dispensing doctors prescribed significantly lower quantities (28). It was discussed whether 10 indicator drugs are enough to draw these conclusions (26). Other UK figures indicate that dispensing doctors issue 12.9% more prescriptions than non-dispensing doctors (37). Similar findings on costs can be found in studies from the USA and South Africa. A study by the Pharmaceutical Assistance Contact for the Elderly found that the average prescription cost paid to physicians was more than one US$ higher then that paid when dispensed from a pharmacy (11). The South African study found that the average cost per script for dispensing doctors to be 120.35 Rand compared to 140.26 Rand for non-dispensing doctors which makes a difference of 3-4 US$ (65b). A UK study evaluated the effect of fundholding in 3 dispensing doctors' practices compared to 5 non-dispensing doctors' practices. All practices were fundholders in the study period from 1990 to 1994. Dispensing doctors were initially found to spend GBP 15.51 per unit ( prescription) versus GBP 15.23 for non-dispensing doctors. By 1994 this increased to GBP 20.48 for dispensing doctors compared to GBP 21.34 for non-dispensing doctors. This development was explained by dispensing doctors being better informed on prescribing costs. Moreover, it could be seen that the savings achieved from fundholding was much smaller than the increase in drug prices, which took place in the study period (18). Also in this study the sample size is very low. Another study on the annual cost of items dispensed by doctors and pharmacists from 1985 to 1991 found a savings in the UK of 23 pence per item when dispensed by dispensing doctors (33). Discussions on this study argue that greater savings would have been found in pharmacies had other factors been included, e.g. discounts, remuneration and fees for containers (32 a, 32b). The cost of prepackaged drugs, in some studies, was found to be higher than pharmacy retail prices. A USA study conducted by the National Association of Chain Drug Stores comparing prices of 50 prescription drugs between the chain drug stores and dispensing doctors, showed that dispensing doctors' prices were on average 13.2% higher (11). Another study compared 20 prescription drugs of given dosage and quantity sold by prepackagers and retail pharmacies. It found in 30% of pharmacy retail prices were lower than the prepackers' prices to the dispensing doctors. When comparing estimated drug costs to be paid by the patient, it was evaluated that 80% of the time, dispensing doctors' prices would be higher than pharmacy prices (59). The American Pharmaceutical Association hired a consulting firm to conduct an economic and policy analysis including a study involving 7600 prescriptions. The use of drugs and costs were compared for physicians with and without ownership interests in repackaging firms. Patients of dispensing doctors with ownership were found to pay 1/3 more for their medication (mean daily medication cost). Prescriptions per patient were higher and own brands were more prescribed. Penicillin was prescribed 8 times more frequently for the owner group. The two groups were compared before ownership and no significant differences in prescribing behavior could be discerned (59). Other studies have found dispensing doctors' prices to be lower. A study by the National Association for Ambulatory Care found that the centers charged on average US$ 4.00 per prescription compared to the pharmacy charge of US$ 13.00 (11). Newsweek conducted a survey on the cost of ampicillin and found it to be obtainable from dispensing doctors for US$ 5, whereas the pharmacy price was from US$ 11.30 to US$ 14.05 (12). In countries where big scale procurement is practiced by chain pharmacies, one would expect pharmacies to be able to obtain more favorable prices, when compared to dispensing doctors, as they can benefit from mass purchasing power. In other countries where competition among pharmacies is intense and patients can shop around one can also expect low pharmacy prices. Other countries have pricing policies and drug prices are controlled, fixed and regulated based on accounts from pharmacies or other parameters. A study from USA comparing 13 dispensing doctors and 38 doctors with intent to dispense with 310 non-dispensing doctors found that doctors positive to dispensing had a lower annual income compared to non-dispensing doctors, 35% versus 20%, earning less than US$ 70,000 per year (30). Figures from UK find that dispensing doctors have a 25.2 % higher annual income compared to non-dispensing doctors (37). Prepacker companies advertise that by dispensing 10 prescriptions a day, dispensing doctors can increase their annual income by US $20,000-61,000 (11,14). The number of prepackaging companies increased from 1987 to 1989 from 3 to 23 (11), indicating an increase in requests for prepackaged drugs and in dispensing practice by doctors in the USA. A Newsweek article estimated dispensing by doctors to account for 0.1% of the total prescribed amount of about US$ 20 billion (11). Furthermore, some authors would argue that the use of prepackaged drugs in safety-sealed containers, packed under GMP, are much safer than drugs taken from larger containers and packed by pharmacists (12). Cost in relation to society was evaluated in South Africa. The average number of visits for dispensing doctors' patients was found to be 35% higher than for non-dispensing doctors. The additional number of visits for medical aid patients has been estimated at 112 million annually. Moreover, an additional 72 million items have been prescribed (65b). Dispensing doctors in the UK prescribed 15% more drugs annually than non-dispensing doctors (28). A deduction scale exists in the UK whereby 9% of the drug bill is returned from the pharmacies to the government. This has been compared with 6% recovery from dispensing doctors. However, the author points out that it must be considered that dispensing doctors negotiate better and obtain lower prices from companies due to the fact they prescribe the medicines (37). All drugs prescribed must, however, be available at all times, and any system must ensure this overall aim. Moreover the drugs must be affordable. Any system which allows dispensing doctors to practice must ensure the availability of the high selling drugs as well as low turn over, low profit drugs at an affordable cost. A pharmacy, in order to be viable and make up the lost revenue from the doctor-dispensed drugs, must charge higher prices on the other drugs it dispenses. This might result in an overall increase in drug costs to the consumer (15). A pharmacy that has lost profits from high turn over drugs will have difficulties in staying viable, leading to reduced accessibility. The increase in dispensing by doctors in the UK has been accompanied by an increase in the number of over-the-counter drugs. This may be a way whereby pharmacies can increase their profit and still be viable in spite of their loss related to increased dispensing by doctors. The benefits of this development have still to be evaluated (74). Conclusion Based on the reviewed literature, it can be concluded that actual studies comparing the practices of dispensing doctors and non-dispensing doctors is very limited. Studies comparing dispensing practices of doctors compared to pharmacists is also very limited and often related to emotions, beliefs and personal opinions. The large number of questions raised by various authors which remain unanswered indicates the importance of this subject. Based on the study findings some conclusions can be drawn. However these findings need to be reevaluated in view of the method applied, e.g. the low number of dispensing doctors included for comparison, or national differences or development of the country. The findings are all closely linked to the laws regulating dispensing by doctors. This must be considered in evaluating the results. Furthermore, most of the studies and discussion have taken place in English speaking countries, especially the UK and the USA, and in developing counties such as South Africa and Pakistan. Comparing non-dispensing doctors to dispensing doctors or their practices, in view of these reservations, the following can be concluded. 1. Background factors: dispensing doctors' patients are normally older, with a higher proportion of them being over 65 years old; dispensing doctors have less years in practice; dispensing doctors have a less positive attitude to pharmacists; when the law allows dispensing by doctors in relation to rurality, more dispensing doctors will be situated in rural areas compared to non-dispensing doctors; if the laws do not relate dispensing doctors to rurality, this difference could not be found; consultation time was found to be 1/3 for dispensing doctors in comparison to non-dispensing doctors. Also history taking was found to be less elaborate for dispensing doctors. Only 11% of dispensing doctors provided information to patients on their diagnosis, compared to 37% for non-dispensing doctors. In this study the dispensing doctors were paid through the sale of drugs, whereas non-dispensing doctors were pediatricians who were paid a consultation fee. 2. Rational drug use: the number of drugs prescribed per patient annually and per encounter were found to be higher for dispensing doctors in two studies and the same in another; use of injections by dispensing doctors per encounter was found to be higher for dispensing doctors than for non-dispensing pediatricians; dispensing doctors were found to prescribe more antibacterials, anti-diarrhoeals, and home made mixtures of unknown composition than ND-pediatricians; fewer dispensing doctors prescribe ORS compared to non-dispensing pediatricians; generic prescribing in two studies was found to be lower for dispensing doctors; in spite of financial incentives (fundholding) it did not reach the level of non-dispensing doctors; in two studies, the average number of visits per patient per year was found to be higher for dispensing doctors' patients, who were also issued 13% more prescriptions. 3. Economics: annual cost per patient from dispensing doctors practices was, in several studies, found to be higher (10 -13%) than for non-dispensing doctors; dispensing doctors were found to prescribe lower quantities per encounter; dispensing doctors may be better informed on drug prices than non-dispensing doctors; one study found dispensing doctors to have a lower income than non-dispensing doctors and another found dispensing doctors to have a higher income. On evaluating good pharmacy practices by dispensing doctors and comparing their practices with those of pharmacies, the following conclusions can be drawn, taking into consideration the above mentioned reservations: 4. Interactions and compliance: no significant differences were found between patients admitted to hospital by dispensing doctors and by pharmacies in regard to adverse events, non-compliance and mis-compliance, evaluating about 4500 admissions. 5. Dispensing and storage based on Ministry of Health inspections in South Africa: 52% of dispensing doctors practices dispensed medicines by inappropriately trained persons; 33% of dispensing doctors practices counted tablets by hand; 61% of dispensing doctors practices had unsuitable containers; labeling was unsatisfactory in 27% of the dispensing doctors practices; storage conditions were inappropriate in 46% of the dispensing doctors practices, with unsatisfactory cool storage facilities in 39%, no expiry monitoring system in 30% and medicines not clearly labeled in 20%. 6. Economic factors: refunds for average prescription costs were found to be higher for dispensing doctors than for pharmacies; comparisons of drug sales or procurement prices for prepackagers and pharmacies, in most studies, found that dispensing doctors' prices were more expensive for the majority of the drugs evaluated than those sold by pharmacies. Recommendations Based on this review it is not possible to recommend one best solution for dispensing drugs whereby the aims of a national drug policy can be ensured. Accessibility is important for many patients in our often very busy society, but whether this should best be achieved by having drugs dispensed by the prescribing doctor or by satellite pharmacies is not clear. However it can be concluded that: a. to safeguard the patient, dispensing must be regulated and the regulation implemented; b. much more knowledge about the benefits and costs of various solutions must be addressed, if appropriate policies and laws are to be found; c. researchers should be urged to increase study on dispensing from the perspective of society and how the aims of availability, accessibility, affordability, safety and rational drug use can best be ensured. Development The number of dispensing doctors has recently been seen to increase in several countries. In the UK about 12.5% of general practitioners dispensed drugs in 1966/67. This increased slightly to 14 % in 1989-90 and to 15% in 1993. Dispensing doctors account for about 7% of all primary care prescriptions. In the USA about 40% of all physicians dispensed in 1923, which decreased to 25% in 1947; 8.7-10% in 1967; and about 5%-8.5% in 1990-91. Other US authors believe that the number of dispensing doctors will continue to increase mainly because of the increase in factories producing prepacked drugs. Other reasons given are to benefit patients and society, and because dispensing offers the possibility to increase physicians' declining income (due to a decrease in patient load). In South Africa the number of retail pharmacies have remained almost constant from 1985 to 1995, whereas the number of registered dispensing doctors has doubled from about 4000 to almost 9000 in the same 10 years. Quality of Care The heated debate on the practice of dispensing doctors is being waged over a complex variety of ethical, economic, public- health and patient care issues: conflict of interest information price control freedom of choice rational drug use convenience ethics $ Actual studies on dispensing doctors are very limited and much of the available information is anecdotal. Only a few of the studies have been repeated and several of the findings are based on very small sample sizes. Factors influencing prescribing of dispensing doctors Various factors have been found to influence the prescribing process and determine its rationale, which include: education, diagnostic and therapeutic skills pharmacological knowledge advice from colleges habits of the "copied prescribers" use of formularies or standard treatment guidelines years of practice/ education personal clinical experience demands by patients prescribers personalities marketing activities from the drug industry consultation time importance of maintaining patient load quality assurance control and regulatory measures information and knowledge on drug costs. It is difficult to distinguish between the effects of individual factors, and little information is available on their actual effects. Only very few studies have evaluated dispensing practice carried out by doctors in relation to these factors. Summary of rational drug use findings: These findings indicate that dispensing doctors tend to prescribe irrationally with a inclination towards polypharmacy, overuse of injections and brand name drugs, irrational choice of drugs and poorer quality of service. Findings - Dispensing and storage Many laws are ambiguous regarding a dispensing doctor's authority to delegate dispensing. Delegation by pharmacists is generally regulated, whereby pharmacists must personally supervise the activity. Dispensing Doctors Dispensing 52% use inappropriately trained dispensers 20% of the medicines were not clearly labelled patient information was unsatisfactory in general 11% of patients have had communication about their condition 33 % counted tablets by hand 61% did not have suitable containers Storage 46% with unsatisfactory storage conditions 70% with no expiry date monitoring system Health economic factors When calculating costs in relation to dispensing by doctors, no model for comparison or evaluation of affordability exists. Procurement cost, sale cost, patient cost (with or without fees, subsidies, taxes etc), net costs, health insurance schemes costs, or cost for the society applied annually or per encounter, have all been applied in evaluating dispensing by doctors. Findings confirmed in more than one study are: Annual cost per patient 10-13% higher for dispensing doctors (SA,UK) Dispensing doctors prescribe lower quantities per encounter and cost is found 5-17% less for dispensing doctors (UK, SA) Patients of dispensing doctors have, per year, 35% more visits and are prescribed 13 % more prescriptions (SA,UK).