Health Policy and Planning Advance Access published online on September 26, 2007
Health Policy and Planning, doi:10.1093/heapol/czm035
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Adverse pharmaceutical payment incentives and providers’ behaviour: the emergence of GP-owned gateway pharmacies in Taiwan
1Associate Professor, Institute of Health and Welfare Policy, and Research Center of Health and Welfare Policy, College of Medicine, National Yang-Ming University, Taiwan.
2Assistant Professor, Institute of Health Services Management, College of Public Health, China Medical University, Taiwan.
3PhD candidate, Institute of Public Health, College of Medicine, National Yang-Ming University, Taiwan.
*Corresponding author. No.155, Sec. 2, Linong St., Taipei City 112, Taiwan, R.O.C. Tel: +886-2–2826 7055. Fax: +886-2–2823 6378. E-mail: yclee{at}ym.edu.tw
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResultsDiscussion and policy...ConclusionReferences |
|---|
Objectives We investigated whether a 2002 pharmaceutical payment reform policy, which provided adverse incentives, fostered an increase in market share of ‘gateway pharmacies’ (G-pharmacies—pharmacies owned, operated and located by the same clinics that prescribe medicines); what the financial impact of G-pharmacies to the clinics is; and what factors determine whether a clinic decides to open a G-pharmacy.
Methods Using the database of the National Health Research Institutes, we collected secondary data on all of Taiwan's National Health Insurance prescription claims from pharmacies and clinics between 1997 and 2003. A G-pharmacy was defined as a pharmacy in which more than 70% of the prescriptions it filled came from the same clinic, which prescribed at least 900 prescriptions monthly, more than 70% of which were released to the pharmacy. Trend plot and frequency were used to analyse the distribution of G-pharmacy data. Logistic regression was used to explore what factors determined whether a clinic decided to open a G-pharmacy.
Results After the 2002 reform, the percentage of total prescriptions filled by G-pharmacies reached 78.71%, the increase in percentage (15.23%) was the highest ever and significant (P < 0.01). The reform's adverse payment incentives resulted in a loss of NT$1.86 billion New Taiwan dollars to all clinics and resulted in a reduction in Taiwan's 2003 fee schedules under the global budget payment system. The decision to establish a G-pharmacy was associated with a clinic's being located in less urbanized areas, being a group practice, having higher patient volumes, being a general practitioner, and being privately owned.
Conclusion The 2002 reform's adverse incentive fostered a significant increase in the market share of G-pharmacies, and reduced the earnings of clinics which did not own them. It is necessary to break the link between profits from pharmaceutical sales and physician prescribing behaviour to prevent the conflict of interest in how medicines are prescribed.
Key Words: Provider payment, pharmaceutical costs, drug prescribing, dispensing policy, gateway pharmacy, national health insurance
KEY MESSAGES
|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussion and policy...ConclusionReferences |
|---|
Financial incentives can be used to improve the quality of medical care (IOM 2001
), or, if used unwisely, can decrease that quality. In Taiwan, one controversy is whether physicians should be allowed to profit from the medications they prescribe. This underlying conflict of interest between economic and clinical benefit might bias prescribing decisions. In order to reduce this possibility, to improve the appropriateness of drug use and to better ensure the autonomy of pharmacists, many Asian countries have implemented a Separation of Prescribing and Dispensing (SPD) Policy (Hsiao 1991; Chou et al. 2003; Kwon 2003). Previous studies have investigated the impact of SPD on providers’ behaviour or on the control of pharmaceutical costs (Chou et al. 2003; Kwon 2003; Huang et al. 2004), but few have evaluated how changing financial incentives to control profits made by prescribing medicine might affect practice behaviours of physicians. We attempt to address this issue in this study.
The health care system in Taiwan is characterized as a closed-staff system, in which most physicians are either practicing independently in clinics or are employees of hospitals. As is the case in many Asian countries (Rodwin and Okamoto 2000; Chou et al. 2003; Kwon 2003), physicians who practiced at clinics in Taiwan both prescribed and dispensed drugs before its SPD Policy was implemented in 1997 (Hsiao 1991; Chou et al. 2003). Another characteristic of Taiwan's health care system is its National Health Insurance (NHI) system, which was reported in 2006 to cover the medical care of 99% of its population. It covers most medical services, prescription drugs and medical devices on a fee-for-service basis (Chiang 1997; Cheng 2003). The Bureau of National Health Insurance, aware of the conflict of interest that might arise from a physician's prescription and sale of medications, has set a flat payment rate per day (NT$35 per day; US$1 = NT$32) to cover the costs of pharmaceuticals for 3 days or less prescribed by physicians at clinics. Physicians can also file itemized claims, which are subject to utilization reviews that may result in total reimbursement, denied reimbursement or reduced reimbursement as well as penalties (Chou et al. 2003; Huang and Chiang 2004).
This flat payment rate is similar to a ‘payment per case’ method of reimbursement but without case-mix adjustment. It allows physicians to keep savings from the provision of pharmaceuticals (Lee et al. 2006), and to avoid reductions in reimbursement brought on by a utilization review. Therefore, most physicians (76.62% in 2000) choose to be reimbursed on a flat payment rate (Huang et al. 2004). This flat payment rate is a means of controlling and freezing prices, which, according to international experience, might have limited effects in controlling total expenditures because physicians tend to make up their losses by inducing the demand for higher volumes of services (Evans 1974; Rice 1983; Reinhardt 1985; McGuire and Pauly 1991; Chou et al. 2003). For example, physicians may un-bundle their services by prescribing fewer days of prescriptions (2–3 days) and encouraging patients to return for visits that might not be otherwise necessary. Moreover, when the amount of the flat payment rate is high, physicians have adverse financial incentives for unnecessary prescriptions. In 1997, drugs were prescribed in 97% of outpatient visits in the clinic sector (Chou et al. 2003). Even after Taiwan's Department of Health phased in the SPD policy over 4 years starting in March 1997, this behaviour of prescribing medications at almost all outpatient visits continued (Huang et al. 2004; Chou et al. 2003). The SPD policy requires that all prescriptions be dispensed by a pharmacist, except for those prescribed in rural areas or under emergency conditions. This requirement should improve the quality of pharmaceutical services theoretically. However, under pressure from the Taiwan Medical Association, the Department of Health conceded to allowing physicians to hire pharmacists to dispense drugs at their clinics (on-site pharmacist), which would allow them to keep the profits that an outside pharmacy would make. Therefore, while the addition of an on-site pharmacist might improve the pharmaceutical services, the fact that the person who prescribes the medications also owns the pharmacy that fills the prescription means that the conflict of interest is not resolved.
In order to reduce the possibility that this incongruous financial incentive would exacerbate conflict of interest, the national insurance system pays a physician a small bonus, a prescription release fee of NT$25, if he or she lets an independent community pharmacy (referred to hereafter as community pharmacy) fill the prescription. Because the bonus is small, few clinics release their prescriptions to community pharmacies to fill (8.27% in 2000, Huang et al. 2004). Whether a clinic decides to release them to a community pharmacy has been associated with degree of urbanization and supply of physicians in townships where a clinic is located, patient volume (visits), type of medical specialties, and type of services which a clinic provides (Huang et al. 2004). One study, conducted by Chou et al. (2003) on the impact of SPD policy in Taiwan, has suggested that although SPD might effectively reduce prescriptions and reduce drug expenditures, it might not reduce total expenditures, particularly in clinics that have on-site pharmacists.
The Global Budget Payment Initiative, which sets a cap on the total health care expenditures of clinics, became effective for clinics in July 2001. This global budget was implemented to improve the quality of primary care and make better use of health care resources. It made it possible for the Bureau of National Health Insurance, with the support of the Taiwan Medical Association, to reduce the flat payment rate for the reimbursement of medication from NT$35 to NT$25 per day from January 2002 (Table 1), further ameliorating the adverse effects of the SPD pharmaceutical incentive. This reduction significantly reduced total drug expenditures (Lee et al. 2006), allowing the Bureau to raise the consultation fees of physicians with the hope that it would encourage the release of more prescriptions to community pharmacies.
|
There was a loophole, however, which brought about a partial failure of this new policy. While the flat payment rate for clinics had been adjusted, that for pharmacies (NT$30 per day) had not, thus creating an undesirable incentive for the clinics which had hired pharmacists to establish so called ‘gateway pharmacies’ (G-pharmacies) so that they could make more of a profit. A G-pharmacy, called a ‘front-door pharmacy’ in Taiwan, is owned and operated by the physicians of a clinic, who usually locate the G-pharmacy at the gateway (front door) or next-door to their clinic. These G-pharmacies allow physicians to earn a bonus prescription release fee (NT$25) for releasing prescriptions (even to their own pharmacies), as well as to profit from the discrepancy in the flat payment rate (NT$5 per day) and a pharmacy's dispensing fees (NT$11) paid by Bureau of National Health Insurance. This loophole makes it possible for physicians in these clinics to earn an additional NT$51 ($25 + $5 x 3 days + $11) beyond their regular National Health Insurance reimbursements (consultation fees). All this and they still enjoy the profits from the already reduced flat payment (Table 1). The loophole allows physicians who have established their own pharmacies to earn an average of at least 13% more per visit (NT$51/NT$382, the average cost per visit in 2002) (NHI-HENC 2005
).
As a result of this new policy, the number of G-pharmacies has increased rapidly. The financial impact on total health care expenditures is also substantial. However, based on the global budget, the more services or drugs that physicians prescribe, the smaller the conversion factor (cf) or the dollar value for each service listed on the NHI's Fee Schedules, a schedule that uses a point system to quantify the relative input of each medical service. Cf is calculated as follows:
|
|
), possibly a result of the establishment of G-pharmacies. The objectives of this study were to understand the prevalence and financial impact of G-pharmacies on all clinics and to explore the factors associated with the establishment of G-pharmacies. The specific research questions were whether the 2002 pharmaceutical payment reform policy significantly fostered the increase in the market share of prescriptions dispensed by G-pharmacies, what financial impact G-pharmacies have had on clinics, and what factors determine whether a clinic owner decides to open a G-pharmacy.
|
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussion and policy...ConclusionReferences |
|---|
The study subjects consisted of all the pharmacies and the clinics that have service agreements with Taiwan's Bureau of National Health Insurance. We conducted secondary data analyses of National Health Insurance claim and prescription databases that we accessed through the National Health Research Institutes (NHRI).
To define gateway pharmacy, we first developed two measures calculated from the database: (1) prescription concentration rate of the clinic (PCRc) and (2) prescription concentration rate of the pharmacy (PCRp). PCRc and PCRp were defined as follows:
|
|
To define G-pharmacy using the above definition, we assumed that a pharmacy would be more likely to be a G-pharmacy established by clinic i if clinic i released most of its prescriptions to pharmacy p (high PCRc) and most prescriptions dispensed by pharmacy p were prescribed by clinic i (high PCRp). An additional criterion was that their total number of prescriptions should be high enough to be commensurate with the incremental cost of hiring a pharmacist. Thus, we arbitrarily defined a pharmacy as a G-pharmacy if nPc of the dominant clinic i was greater than 900, and the release rate of prescriptions of clinic i was greater than 70%; and both PCRc and PCRp were greater than 70%. In the subsequent sensitivity analysis, we replaced the above parameters (threshold on PCRc, PCRp and release rate) with 50%, 60%, and 80%. As these proxies did not yield substantially different results, we reported our results using the original thresholds only.
To explore the clinic characteristics that might determine the likelihood of a clinic establishing a G-pharmacy, we created five variables: (1) Type of clinic registered by the physician—general practitioner vs. specialists; (2) Ownership of clinic—public vs. private; (3) Mode of practice—solo (1 doctor), paired (2 doctors) and group (3 or more doctors); (4) Daily service volume—the mean daily number of clinic patients divided into five groups (<31, 31–50, 51–70, 70–120, >120 visits); and (5) Urbanization level of clinic location—the cities/villages/districts were grouped into eight levels (1–8) according to Tzeng and Wu (1986), lower numbers here representing higher urbanization levels; to simplify the analysis, the levels were further grouped into three categories: high (1–3), medium (4–5) and low (6–8).
Most data about the clinics were compiled from four databases:
- From the Expenditures for Prescriptions Dispensed at Contracted Pharmacies, we could obtain data on the time, numbers and source of prescription dispensed by the pharmacy. Each prescription could be traced by the identification number of the prescribing physician (clinic). The dominant clinic was defined as the clinic which contributed the highest number of prescriptions to that pharmacy.
- From the Registry for Contracted Medical Facilities, we obtained information on the type, ownership and location of a clinic.
- From the Registry for Medical Personnel, we obtained physician identification numbers (to define mode of practice) and those of pharmacies in a clinic.
- From the Monthly Claim Summary for Ambulatory Care Claims (CT), we collected data on the time and volume of outpatient visits.
All databases were checked and, if necessary, cleaned and merged. To define G-pharmacies, database 1 was compiled by pharmacy identification number and then merged with database 4 to determine the dominant clinic and to calculate nPp, nPpi, nPc, PCRc and PCRp. The above data were then merged with database 2 (facility) and database 3 (personnel) to obtain the measures of clinic characteristics.
Descriptive statistics were used to describe the distribution of G-pharmacy and clinic characteristics. Logistic regression was used to analyse the relationship between clinic characteristics and likelihood (odds ratio, OR) that a clinic would establish a G-pharmacy. All statistical operations were performed using SAS® statistical software.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussion and policy...ConclusionReferences |
|---|
We collected 2 068 195 claims for prescriptions dispensed at contracted pharmacies, including community and G-pharmacies in 1997, 5 347 234 in 1998, 10 024 655 in 1999, 12 481 682 in 2000, 15 306 865 in 2001, 31 190 019 in 2002 and 45 918 393 in 2003. Organizing the data by pharmacy identification number, we found 3559 pharmacies by the end of December 2003.
As can be seen in Figure 1, in January 1997 there were about 2665 pharmacies. The number peaked in 1998 to around 3400, reduced gradually between 1999 and 2001, and rose again after 2002. Table 2 shows that the number of pharmacies decreased by between 3–6% each year between 1999 and 2001. It rose by 17.06% and 6.30% in 2002 and 2003, after the flat payment rate was reduced (January 2002). The number of clinics remained about the same between 1998 and 2001, but increased in 2002 and 2003. Figure 2 shows that in January 1997, there were only six G-pharmacies. Table 2 and Figure 2 indicate that before the reform, the number of these pharmacies increased by a higher rate during 1998–99 (179.17 and 76.12% increase, respectively) and by a lower but still high rate during 2000–01 (21.61 and 32.75% increase, respectively). After the reform, the number of G-pharmacies rose significantly again (P < 0.01), by 128.87% in 2002 and 38.19% in 2003.
|
|
|
The proportion of pharmacies dispensing any prescriptions during a specific period experienced the biggest increase during 1997–98 (18.48%) and a modest increase during 2000–03 (6–10%). There was only a marginal increase in 1999. It finally reached 77.24% in 2003. G-pharmacies made up only a very small percentage (under 10%) of the total number of pharmacies before 2000. The percentage almost doubled (to 26.05%, from 13.32% in 2001) after the flat payment rate was reduced in 2002, reaching 33.86% in 2003. In 1997, only 0.60% of the clinics owned G-pharmacies. The percentage increased marginally between 1997 and 2001. After the flat payment rate was reduced, the percentage of clinics that owned G-pharmacies grew significantly to 10.63% in 2002 (P < 0.01) and insignificantly to 14.08% in 2003.
Regarding percentage of total prescriptions (market share) prescribed by clinics and dispensed by pharmacies, G-pharmacies filled 38.38% of all prescriptions dispensed by contracted pharmacies in 1997. Their market share increased 5–8% yearly during 1997–2001. The change was even more profound (15.23%) and significant (P < 0.01) in 2002 after the reform, though growth slowed to 0.62% in 2003. G-pharmacies filled 79.33% of all the prescriptions dispensed by pharmacies in 2003. The other 20.67% of the prescriptions were filled by community pharmacies. Overall, there was little change in these measures through 2001, but once the flat payment rate was reduced in 2002, they all increased substantially.
Focusing on 2003 data, we further examined the financial impact of G-pharmacies on total expenditures. There were 45 918 393 prescriptions prescribed by clinics and dispensed by pharmacies in 2003; of these, 36 427 330 were dispensed by G-pharmacies. As mentioned earlier, clinics with G-pharmacies could make an extra 51 New Taiwan dollars per prescription, resulting in extra billings of NT$1.86 billion (51 x 36 427 330) to the Bureau of National Health Insurance in 2003. In other words, clinics which set up G-pharmacies on average had billings amounting to NT$1 540 000 more than they should in 2003. This was taken from a pre-set global budget, and the extra expenditure of NT$1.86 billion was in fact shared by all the clinics, including those which had not opened G-pharmacies, because the conversion factor of the NHI Fee Schedules for medical services was reduced.
Based on logistic regression, we found the following five kinds of clinics to be significantly (P < 0.01) more likely to establish G-pharmacies: (1) clinics registered as general practitioners vs. specialists [OR = 1.23 (1.07–1.41)]; (2) clinics with three or more physicians [OR = 2.76 (2.31–3.21) vs. solo practices] and clinics with two physicians [OR = 1.68 (1.42–2.00) vs. solo practices]; (3) clinics with higher numbers of daily ambulatory visits (31+); (4) privately owned clinics vs. public clinics [OR = 7.66 (3.33–17.66)]; and (5) clinics located in areas with the lowest levels of urbanization vs. those in areas with the highest [OR = 1.70 (1.41–2.05)] (Table 3). However, no such statistical difference was found between those located in areas with medium levels of urbanization vs. those in areas with the highest [OR = 1.16 (0.99–1.36)].
|
|
Discussion and policy implications |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussion and policy...ConclusionReferences |
|---|
In this study, we evaluated the impact of a 2002 pharmaceutical payment reform policy on the growth and financial impact of G-pharmacies. Our first major finding was that although the number of G-pharmacies has been increasing since 1997, it was not until 2002 when the flat payment rate for medications was reduced that great and significant increases (15.23%) occurred in the percentage of total prescriptions filled. Based on these findings, we believe that the reduction in the flat payment rate for medications might foster an increase in the percentage of total prescriptions filled by G-pharmacies, filling the pockets of G-pharmacy owners (average NT$1 540 000 per clinic per year) and lowering the total earnings of clinics by NT$1.86 billion, which is particularly unfair to those clinics which have not established G-pharmacies.
Clinics that own G-pharmacies or hire pharmacists to work on-site both provide convenient and quick dispensing services, and usually have no problems in the availability of pharmaceuticals prescribed by their physicians, but those that release their prescriptions to community pharmacies may not have these advantages. Dispensing medication on-site under the name of the clinic or through G-pharmacies can present a conflict of interest, because the physicians standing to profit from the sale of medications may be tempted to prescribe drugs their patients do not need. Because the pharmacists in both cases are employees of the clinics, they might not work independently from their employers, or they might not perform the role of ‘checking and balancing’ between the two professions that the SPD policy aims to separate.
Based on the above discussion, why are an increasing number of physicians establishing G-pharmacies rather than employing pharmacists at their clinics in Taiwan? First, physicians who establish G-pharmacies benefit from savings on income taxes, because if the profits from pharmaceutical sales are not taken into account, their rate of income taxes may be lower. Secondly, in Taiwan these physicians do not pay the social insurance premiums or contribute to pension plans for the pharmacists working in their G-pharmacies, though ethically they should. Thirdly, they might claim prescription release fees, though by law they should not as this behaviour, although profitable, is not in line with the spirit of the SPD policy. Above these incentives, the reduction of the flat payment rate in 2002 provided further incentive (an extra NT$25 per patient per visit) for clinics to establish G-pharmacies, explaining the significant growth in G-pharmacies in 2002 (Table 1). Similar to the ‘profit-maximizing model’ used to describe hospital behaviour (Feldstein 1988), some physicians in Taiwan run their clinics like hospitals and try their best to maximize their profits by minimizing their costs or increasing their billings.
Our second major finding was that the financial impact of the increase in number of G-pharmacies and the increase in prescriptions filled by them was substantial. They were able to bill an extra 1.86 billion New Taiwan dollars in 2003, 2.29% of the annual global budget allocated to all of Taiwan's clinics by the National Health Insurance programme in 2003, or 81.58% of their total annual budget increase for that sector, which was about NT$2.28 billion per year. This left little room to meet the needs of growing demand or increasing practice costs. Moreover, because all of Taiwan's clinics must share the global budget, the financial consequences caused by a few clinics with G-pharmacies (14.08%) must be borne by all clinics. All physicians were collectively penalized by the dramatic decreases in floating conversion factors of the National Health Insurance Fee Schedules, dropping from 1.16 in 2001 to 1.01 in 2002, 0.98 in 2003 and 0.83 in 2004 (NHI-HENC 2005). Specialty clinics (e.g. general surgeons or ophthalmologists) lost more than clinics operated by general practitioners because what they earned was based less on pharmaceuticals (cf = 1, fixed) than on procedures (subject to floating cf). This difference has resulted in great conflicts among member physicians of the Taiwan Medical Association. Therefore, an incentive comes with benefits as well as pitfalls, especially when there are oversights in incentive design possibly resulting in undesirable consequences. Clinics without G-pharmacies are not the only stakeholders opposed to G-pharmacies. Community pharmacies, represented by the Taiwan Pharmacists Association and faced with losses in market share to G-pharmacies, have been strongly opposed to the establishment of G-pharmacies.
Our third major finding was that the likelihood of a clinic establishing a G-pharmacy was significantly determined by: whether it was a general practitioner or a specialist clinic, how many physicians practiced there, how many prescriptions were being prescribed daily, whether it was privately or publicly owned, and how urbanized its location was. The relationship between having a G-pharmacy and number of physicians and volume of prescriptions is consistent with economic theory; for higher volume clinics and pharmacies, the profit made from a G-pharmacy outweighs the costs of establishing one, more so than it would for clinics with one physician or with fewer prescriptions per day. Besides, the productivities of a group of physicians are more likely to be enhanced by hiring other professionals (Reinhardt 1972). The above findings are also generally consistent with one previous study which found clinics with lower patient volumes and fewer numbers of physicians to be more likely to release prescriptions to community pharmacies (Huang et al. 2004).
Public clinics were less likely to own G-pharmacies, probably because there is less focus on profit-making. The clinics registered as specialists, which relied more on other procedures (e.g. surgical, rehabilitative, laser or other procedures and tests) than pharmaceuticals, prescribed fewer medications per day, and thus were less likely to own G-pharmacies. They have also been found more likely to release prescriptions to community pharmacies (Huang et al. 2004).
It is more difficult to explain why clinics in the least urbanized areas were more likely to open G-pharmacies than those in the most urbanized areas. One probable reason is that community pharmacies were located further away, so having a G-pharmacy would be more convenient for their patients. Another reason may be that they might be seeing fewer patients, and may, therefore, need to make up for the loss in service fees with profits from the sale of medicines. The above findings are only partially consistent with one previous study, in 2000, which found that clinics located in the most and the least urbanized areas were more likely to release prescriptions than clinics located at areas with middle levels of urbanization (Huang et al. 2004). The difference in findings might be because the reference groups and observation time (2003 vs. 2000) were different.
The Bureau of National Health Insurance has tried to mediate the disputes over G-pharmacies between the Taiwan Medical Association and the Taiwan Pharmacists Association since 2004. In the amendment to the 2005 global budget agreement, the National Health Insurance Health Expenditures Negotiation Committee (NHI-HENC), responsible for global budget negotiation and allocation, requested that the problem caused by G-pharmacies be resolved before the end of 2005. In January 2006, the Bureau finally reached agreement with these two associations to reduce the flat payment rate reimbursed to pharmacies to the same level as that reimbursed to clinics (NT$25 per day). The savings will be used to raise the dispensing fees paid to any notarized independent community pharmacies, increasing from NT$25 to NT$45. Despite this equalization measure, although rates have been slightly reduced, there still exist financial incentives for establishing a G-pharmacy. The difference in financial incentives paid to clinics with on-site pharmacists and those with G-pharmacies decreased from NT$51 in 2002 to NT$49 in 2006 (Table 1). To make up for their losses, some of the clinics with G-pharmacies (around 1000) had their G-pharmacies notarized as community pharmacies, then released the prescriptions to their own notarized G-pharmacy and collected the prescription release fees as well, without being aware this would be considered as submitting a fraudulent claim. The local courts, together with the Bureau of National Health Insurance, have been investigating these fraudulent claims, and have asked those who submitted such claims to turn themselves in before the end of June 2006, or possibly face a maximum sentence of 5 years imprisonment.
To reduce the effect of the adverse incentives, the Bureau of National Health Insurance decided to eliminate the prescription release fees from the Fee Schedules in July 2006. Savings will be used to raise the physician consultation fees for 0–25th visits and raise the pharmacy dispensing fees for 0–79th prescriptions at clinics with on-site pharmacists. The dispensing fees for 80–99th prescriptions have not changed. No dispensing fees will be paid beyond the 100th prescriptions. Consultation fees for 26–30th visits will also be reduced. Therefore, for clinics which have up to 25 prescriptions, the incentives for clinics with on-site pharmacists and G-pharmacies will be the same. In addition, physicians will be required to provide prescriptions to all their patients, who can decide themselves where they would like to get their prescriptions filled. The Department of Health hopes that this new policy will ultimately stop the growth of G-pharmacies and enforce the SPD policy. However, as can be seen in Table 1, if the total profits from pharmaceuticals outweigh the costs of hiring a pharmacist, physicians will still choose to hire on-site pharmacists or to own G-pharmacies rather than to release prescriptions to the community pharmacist, even after July 2006. Therefore, as long as physicians’ incomes are linked to pharmaceutical profits, it is unlikely that the SPD policy will be able to successfully reduce this conflict of interest.
Unlike Taiwan, South Korea has taken a different approach by introducing mandatory implementation of SPD in July 2000. The SPD reform in South Korea has succeeded in totally eliminating the drug profits of physicians and has broken the connection between financial incentives and physician prescribing behaviour (Kwon 2003). However, this has been at great cost to society as physicians’ fees were increased as much as 44%. Added to these costs was the impact of physicians prescribing more expensive brand-name drugs after implementation of the SPD policy (Kwon 2003). Taiwan would not experience this cost increase because pharmaceutical reimbursements are constrained by the flat payment rate for most prescriptions.
Comparisons of both countries suggest that for Taiwan to reduce adverse pharmaceutical incentives and to control pharmaceutical costs in the short-run, it should either keep the flat payment rate but gradually reduce the profits that clinics which hire on-site pharmacists can make from them, or it should set flat payment rates based on the case-mix of patients (a pharmaceutical case payment system). In the long run, however, Taiwan should, like Korea, break the connection between the financial incentives and prescribing behaviour of physicians. If not, the conflict of interest will persist.
Limitations
Our study is limited by our use of administrative claims data and arbitrarily defined measures to identify G-pharmacies. Although we have conducted sensitivity analysis by introducing different thresholds on PCRc and PCRp which yielded similar results, our results should still be considered cautiously.
|
Conclusion |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussion and policy...ConclusionReferences |
|---|
In conclusion, we have found that the adverse pharmaceutical payment incentives fostered a significant increase in market share of G-pharmacies, which filled as much as 79.33% of all prescriptions dispensed by pharmacies in 2003. The financial impact of these incentives on all clinics was equivalent to 1.86 billion New Taiwan dollars in 2003. G-pharmacies were more likely to be established by clinics that were registered as general practitioners, had a group of physicians, prescribed more medications per day, were privately owned and were located in areas with a very low level of urbanization. Again, this study was limited by its use of administrative data only. Therefore, the results should be interpreted with caution.
Based on the above analysis, because physicians are strongly influenced by financial incentives, we have the same recommendation for policy makers as Rodwin (2004), that financial incentives be treated like drugs and adjusted when using them with doctors to ensure that they are safe as well as effective. Besides, as long as physicians are allowed to profit from the sales of medications, many will continue to employ pharmacists in their clinics and conflict of interest problems will remain. Therefore, countries considering implementing an SPD policy may want to develop an acceptable and feasible policy to break the link between pharmaceutical profits and the prescribing behaviour of physicians, while keeping pharmaceutical expenditures in control. For Taiwan, one option is to reduce the sales profits on pharmaceuticals and use the savings to gradually raise physicians’ consultation fees. To understand actual prescribing behaviour and cost impacts, we recommend that future studies compare the prescribing patterns of clinics that release their prescriptions, clinics with on-site pharmacists and clinics with G-pharmacies, and then develop models for physicians to work with professionally independent pharmacists.
|
Acknowledgements |
|---|
This study is based in part on data from the National Health Insurance Research Database provided by the Bureau of National Health Insurance, Department of Health and managed by National Health Research Institutes. The authors thank the Bureau of National Health Insurance and National Health Research Institutes for providing the data for our research. In addition, we gratefully acknowledge the valuable comments by two anonymous reviewers.
The interpretation and conclusions contained herein do not represent those of the Bureau of National Health Insurance, Department of Health or National Health Research Institutes.
|
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussion and policy...ConclusionReferences |
|---|
Cheng TM. Taiwan's new national health insurance program: genesis and experience so far. Health Affairs (2003) 22:61–76.
Chiang TL. Taiwan's 1995 health care reform. Health Policy (1997) 39:225–39.[CrossRef][Web of Science][Medline]
Chou YJ, Yip WC, Lee CH, et al. Impact of separating drug prescribing and dispensing on provider behaviour: Taiwan's experience. Health Policy and Planning (2003) 18:316–29.
Evans RG. Supplier-induced demand: some empirical evidence and implications. In: The economics of health and health care—Perlman M, ed. (1974) New York: Wiley Press. 162–73.
Feldstein P. Health care economics, third edition (1988) New York: John Wiley & Sons. 226–27.
Hsiao ML. The planning and prospective of the separation of drug prescribing and dispensing policy in Taiwan. Journal of Pharmacy (Taiwan) (1991) 7:3–6. [in Chinese].
Huang WF, Chiang HC. Pharmaceutical benefit scheme and cost containment in Taiwan's National Health Insurance. Journal of Pharmaceutical Finance, Economics, and Policy (2003) 12:133–49.
Huang KH, Lee YC, Huang YT. Impact of the separation of drug prescribing and dispensing policy on the release rate of prescriptions from clinics. Taiwan Journal of Public Health (2004) 23:469–78. [in Chinese: English abstract].
Institute of Medicine (IOM). Crossing the Quality Chasm: a new health system for the twenty-first century (2001) Washington, DC: National Academies Press.
Kwon S. Pharmaceutical reform and physician strikes in Korea: separation of drug prescribing and dispensing. Social Science and Medicine (2003) 57:529–38.[CrossRef]
Lee YC, Yang MC, Huang YT, et al. Impacts of cost containment strategies on pharmaceutical expenditures of the National Health Insurance in Taiwan, 1996–2003. PharmacoEconomics (2006) 24:891–902.[CrossRef][Web of Science][Medline]
McGuire TG, Pauly M. Physician response to fee changes with multiple payers. Journal of Health Economics (1991) 10:385–410.[CrossRef][Web of Science][Medline]
National Health Insurance Health Expenditure Negotiation Committee (NHI-HENC). Reference Statistics and Indicators Applied to Annual Global Budget Negotiation (2005) Taipei: Department of Health (DOH). [in Chinese].
Reinhardt UE. A production functions for physician services. Review of Economic and Statistics (1972) 54:55–66.[CrossRef]
Reinhardt UE. The compensation of physicians: approaches used in foreign countries. Quality Review Bulletin (1985) 11:366–77.[Medline]
Rice TH. The impact of changing Medicare reimbursement rates on physician-induced demand. Medical Care (1983) 21:803–15.[CrossRef][Web of Science][Medline]
Rodwin MA. Financial incentives for doctors. British Medical Journal (2004) 328:1328–9.
Rodwin MA, Okamoto A. Physicians’ conflicts of interest in Japan and United States: lessons for the United States. Journal of Health Politics, Policy and Law (2000) 25:343–75.[Abstract]
Tzeng GH, Wu TY. Characteristics of urbanization levels in Taiwan districts. Geographical Research (Taiwan) (1986) 12:287–323. [in Chinese: English abstract].
Accepted for publication 9 July 2007.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  C. D. James, J. Peabody, O. Solon, S. Quimbo, and K. Hanson An Unhealthy Public-Private Tension: Pharmacy Ownership, Prescribing, And Spending In The Philippines Health Aff., July 1, 2009; 28(4): 1022 - 1033. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||