Predicting the impact of Medicare Part D implementation on the pharmacy workforce

Article Outline

• Abstract
• 1. Introduction
• 2. Methods
  • 2.1. Study design
  • 2.2. Forecasting
  • 2.3. Univariate forecasting model
  • 2.4. Multivariate forecasting models
  • 2.5. Dependent variable
  • 2.6. Independent variables
• 3. Results
  • 3.1. Univariate models
  • 3.2. Predictors supporting observation
  • 3.3. Predictors opposing
  • 3.4. Multivariate model result
• 4. Discussion
  • 4.1. Limitations
  • 4.2. Better data are needed
• 5. Conclusions
• References
• Copyright

Abstract 

Background

There is currently a shortage of pharmacist manpower, and it is expected to continue into the near future. It is also likely that the implementation of Medicare Part D will further aggravate the shortage by increasing demand, but it is not clear how much impact it will have.

Objective

The purpose of this study was to estimate the impact that the new Medicare drug benefit program will have on pharmacy workforce demand.

Methods

Analysis was conducted using forecasting techniques, which combines traditional statistical theory with both quantitative and qualitative methods. The Aggregated Demand Index (ADI) was designated as the dependent variable. A number of independent variables were selected for their potential to affect the workforce, demand for prescriptions or clinical services, and patient population. Data for the identified variables were collected from a variety of sources. Supply and demand data were analyzed at a national level.

Results

Both historical and univariate forecasts indicated that the demand for pharmacists will continue to exceed the supply of pharmacists. The ADI ratio of pharmacist demand-to-supply has recently leveled off which means that demand and supply are in an equilibrium that falls to the demand side. Consequently, the Medicare Modernization Act (MMA) is not predicted to produce a dramatic increase in prescription volume, which would change the current demand for pharmacists. Multivariate forecasting models were not robust primarily due to the lack of precise predictor variables.

Conclusions

Despite the reliance on preliminary univariate forecasts and imprecise predictor variables, it appears that the increased use of prescriptions due to the MMA Part D will have minimal impact on pharmacist demand.

Keywords: Medicare Part D, Pharmacists, Manpower, Workforce, Forecasting

Back to Article Outline

1. Introduction 

Recent studies of the health care workforce indicate that the current shortage of pharmacists will persist into the future even though there could be as many as 225,000 pharmacists by 2010.¹ Predictions of pharmacist shortages at the national level have been as high as 157,000 by 2020.²

Researchers and planners alike have been working to characterize, understand, and predict future demand for pharmacy services and supply of pharmacists.¹, ² Additionally, data related to current manpower supply are being collected by the Pharmacy Manpower Project. These data are used to derive regional and state-level estimates of the Aggregate Demand Index (ADI), which calculates the demand for pharmacists on a monthly basis using survey data submitted by panelists in various pharmacy practice sites.³, ⁴

Whether the current shortage is a continuation of the shortage that began in the late 1980s or a more recent, separate event is not clear, because the factors contributing to both shortages appear to be the same.¹, ⁴, ⁵ These factors include the number of prescriptions written or filled, number of pharmacy graduates, diversity of pharmacy practice, and number of part-time workers.¹, ², ³, ⁴

There is now another factor to consider. Medicare's new Part D program, implemented by congress as part of the Medicare Modernization Act (MMA) will undoubtedly increase access to prescription medications and clinical pharmacy services for America's increasingly aging population. The geriatric patient population eligible for this new program is more likely to use multiple medications on an ongoing basis and requires more intense medication therapy management, both of which will further amplify the effects of Medicare Part D implementation on demand for pharmacists. In the United States, the average number of prescriptions for a patient aged between 65 and 90 years has increased from approximately 19 in 1996 to over 25 by 2002 according to the Medical Expenditure Panel Survey. However, the real extent to which this new program will aggravate pharmacist shortages is not known. Therefore, the purpose of this study is to examine the effect of Medicare Part D implementation on the predicted pharmacist workforce shortage using a forecasting method.

Back to Article Outline

2. Methods 

2.1. Study design 

This study attempted to develop a forecasting model to estimate the demand for community pharmacists from January 1, 2006 through December 31, 2009 as a result of the MMA Part D using retrospective data from various public sources. Because no human subjects were used, Institutional Review Board approval was not required.

2.2. Forecasting 

This study used standard economic forecasting to estimate future monthly ADI values from September 2005 through August 2009. Forecasting is a method that uses historical and presents data to infer what to expect in the future. Prior to making specific forecasts, it is necessary to develop a specific model to estimate the dynamic relationship between various predictors.⁶ This study used ordinary least squares to determine the relationships between potential predictor variables and ADI estimates. The next step involves developing the forecast model to predict future values based on historical data and predictor variables.

Although forecasting models use traditional statistical (objective) methods, assessment of each model forecast relies on both subjective and objective techniques. The objective measures used to determine the robustness of our models included the R² value, Durbin-Watson statistic, Akaike Criteria (AIC), Schwarz Criteria (SIC), and partial autocorrelation function (PACF).⁶ The subjective criteria used to examine the validity of the forecasted model included how closely the predicted estimates matched historical data and how reasonably the forecasted estimates corresponded to the expected demand.

2.3. Univariate forecasting model 

Both univariate and multivariate models were developed, and all of the models were examined for the presence of a stochastic or deterministic trend, seasonality, and cycles. The univariate model used 1 variable, monthly ADI estimates, and forecasted future estimates (January 1, 2006 through December 31, 2009) solely on the basis of its own past values. Thus, univariate forecasts do not directly consider the influence of the MMA Part D, but provide an understanding of the overall trend in the demand for pharmacists.

2.4. Multivariate forecasting models 

The multivariate models were developed and forecasted on the basis of their own historical data as well as other predictor variables. The multivariate models used both a conditional and an unconditional approach. Essentially, an unconditional model allows the researcher to generate forecasts based on the past performance of each of the predictors using a vector autoregression (VAR). The unconditional forecasts for each of the predictors were subjectively compared to forecasted estimates found within the literature to determine the appropriateness. Historical monthly prescription volume was obtained from National Association of Chain Drug Stores (NACDS) and was used to capture the influence of the MMA Part D on the demand for pharmacists. The historical data for both monthly prescription volume and other predictor variables were used in a VAR to create forecasts of future values. Monthly prescription volume estimates obtained from the literature as a consequence of the MMA Part D were then compared with those derived through VAR. It was not expected that those estimates derived from VAR would closely reflect the prescription volume as a result of the MMA Part D. Thus, VAR predictions were modified as needed to more closely reflect those values predicted in the literature as a result of the MMA Part D.

In contrast, a conditional model requires the researcher to subjectively input future predictor values, which is then used to forecast future ADI estimates. This method allows monthly prescription volume estimates obtained in the literature to be imputed over the forecasted time frame. The usefulness of the predictors in the models was examined using a Granger Causality test.⁶ The historical time frame used to develop the multivariate forecasts was March 2001 through December 2005. This time frame is shorter than that of the univariate forecasts due to the limited availability of monthly prescription volume data.

2.5. Dependent variable 

Of the several measures or estimates of pharmacist workforce reviewed as potential dependent variables, the ADI, which is a ratio of pharmacist demand-to-supply, was selected.³ The ADI appeared to have the most robust mechanism for deriving the severity level of the pharmacist shortage.

As mentioned earlier, the ADI estimate was developed as part of the Pharmacy Manpower project and predicts the nationwide and regional demand for pharmacists in a community, institutional, and combined settings. More specifically, through monthly surveys to a panel of pharmacist employers nationwide the ADI quantifies the difficulty in filling pharmacist vacancies. The ADI estimate is a ratio of demand-to-supply that ranges from 1 to 5 with the lower numbers indicating that the demand is much less than the supply and the higher numbers indicating a high demand with difficulty in filling open positions. A mid-range value of 3 indicates supply is meeting demand.

The current month's ADI estimates of workforce demand are reported by site, state, and region. Because this study needed only those estimates related to community retail dispensing, which would be affected by the MMA Part D implementation, only the national community site estimates were used. The national estimates of manpower demand are based on aggregated individual state-level ADI values and the percentage of the total U.S. population that resides in each of the states, resulting in a weighted ADI average. Thus, the national ADI estimate for community sites is a composite measure addressing the overall demand for pharmacists in the United States that month. The data used to estimate ADI have been captured since August of 1999. This study used monthly ADI estimates from August of 1999 through August of 2005 to develop a forecast model. The actual monthly estimates were obtained directly from Dr Katherine Knapp, one of the individuals instrumental in the development of the ADI.³

2.6. Independent variables 

Several independent variables were identified based on prior research that predicted the severity of the pharmacist shortage and were used in the development of the multivariate models.

Table 1 provides a description of the operationalization and source for each of the independent variables considered in various models.

Table 1. Variables used in forecasting models

AACP, American Association of Colleges of Pharmacy; ACPE, Accreditation Council for Pharmacy Education; BLS, Bureau of Labor Statistics; IMS, Intercontinental Marketing Services; NABP, National Association of Boards of Pharmacy; NACDS, National Association of Chain Drug Stores.

According to the 2000 Health Resources and Services Administration (HRSA) report, the demand for pharmacists is the result of several factors including increased use of prescription medications; market growth of retail pharmacies; expansion of pharmacists' roles; changing demographic profile of pharmacists; and burden placed on pharmacists by third-party payers.⁵ These factors were used to operationalize a number of the predictor variables. The 2000 HRSA report was descriptive and did not quantitatively link the shortage of pharmacists with the variables. Other sources were used to derive quantitative information.

One of the main drivers influencing pharmacists demand is prescription volume and its growth rate as documented by Knapp et al.⁴ Monthly prescription volume data were obtained from NACDS. It is also important to consider, direct-to-consumer (DTC) advertising not only because it influences the total prescription volume, but it also engages the pharmacist-patient relationship. Thus, the annual amount spent on DTC was captured from Intercontinental Marketing Services. Additionally, use of mail-order pharmacies tend to decrease retail prescription volume; thus, mail-order use was included as a predictor variable.

A number of other factors indirectly influence the demand for pharmacists. Clearly, the number of registered pharmacies influences the overall demand. As a result, the annual number of registered pharmacies was considered in the multivariate models. The expanded professional opportunities both within and outside the retail setting are expected to further influence the demand for pharmacists. As a consequence, the percent of all pharmacists who are working within a community setting was considered. Additionally, as the demand for pharmacists and their roles expand, states have increased the maximum number of technicians that pharmacists can supervise. As a consequence, the number of technicians, the ratio of pharmacists to technicians, and mean technician salary were considered in the multivariate analysis.

Third-party prescriptions have been found to have a negative affect on pharmacists' productivity. Research has indicated that between 9.5% and 20% of a pharmacist's day is spent addressing third-party prescription issues.⁵ Thus, the percent of prescription dispensed through a third-party administrator was considered.

The multivariate analysis also examined the number of active pharmacists, new graduates, and retiring pharmacists.¹ Pharmacists' salaries were also included in assessment of supply side factors because their upward trend may influence the overall demand for pharmacists.⁵

Overall, there was limited quantitative research assessing the predictors of pharmacists demand. Walton et al,¹⁵ quantitatively validated the correlation between the number of filled pharmacy positions and the number of new graduates, prescription volume, hospital days, number of licensed community pharmacies, and state residents aged older than 65 years. Regression models found that the state population and community pharmacy prescription volume were the only significant variables in predicting the number of filled pharmacist positions.¹⁵ Clearly, some of the proposed predictor variables have stronger rationale supporting their use in multivariate models than others as it relates to predicting the demand for pharmacists, but it was thought that being more comprehensive may enhance further research. Due to potential overlap of influence among some of the variables, the presence of multicollinearity was also examined.

Unlike the ADI and prescription volume, which are monthly estimates, all of the predictor variables selected for this study are assessed annually. Annual estimates were converted to monthly estimates to maximize the sensitivity of forecasting models. Where possible, additional information was used to account for seasonal variations. When none was available, total amounts were divided into 12 equal parts (eg, pharmacists' wages) and ratios or means were maintained at the same value each month (eg, percentages of mail-order prescriptions or community pharmacists).

Unlike the monthly ADI estimates, which were obtained through August 2005, most of the predictor estimates were only available up to December 2004. As a result, univariate forecasting models were developed to estimate the values for the “missing” months (January 2005 through August 2005) using the criteria explained above to identify the most appropriate forecasts. The annual first professional degree graduate estimates were divided into December graduates (20%) and May graduates (80%) to better indicate when entry-level pharmacists were entering the workforce.

Back to Article Outline

3. Results 

Overall, there was an unmet demand for pharmacists available to fill open positions (Figure 1) within a community setting throughout the historical data (August 1999 to August 2005). The magnitude of the unmet demand has been slowly trending downward from an ADI of 4.0 to the mid-3 over a 5-year period until January 2005 when the ADI leveled off at around 3.5 implying demand, and supply factors were in equilibrium at a point where demand was still greater than supply but the situation was stable.

• 
  • View Large Image
  • Download to PowerPoint

• Fig. 1. 

  Historical and univariate forecasted Aggregated Demand Index estimates from 1999 through 2009.

3.1. Univariate models 

The univariate model was developed based on the following model (see Equation 1):

(1)

β_o = y Intercept; β_time = time period; AR (1-3) = first through third order autoregressive processes; seasonality = seasonal variation; ε_t = unexplained error term.

This model predicted approximately 55% (adjusted R²) of the variance in the ADI. The Durbin-Watson statistic approached 2, which indicated no autocorrelation, and the univariate model was developed to minimize both the AIC and SIC values. The resulting univariate forecast indicated that the demand for pharmacist will continue to decrease. More specifically, the forecasted demand-to-supply ratio (ADI) will approach 3.0, which indicates a sufficient supply of pharmacists given the demand. However, the confidence intervals for the model were large, indicating the potential for significant variation in the forecasted estimates and a need for caution when interpreting the results (Figure 1).

The demand-to-supply ratio for pharmacists was decreasing at the same time the volume of prescriptions (Figure 2) was increasing. The number of yearly prescriptions dispensed increased by 6% between 2002 and 2005. Future increases in prescription volume due to the implementation of MMA Part D were estimated by Heffler et al to grow by 8.9% for 2005-2006 and 6.5% for 2006-2011, which indicates continued increases in prescription volume.¹⁶ Tests of individual predictor variables produced mixed results.

• 
  • View Large Image
  • Download to PowerPoint

• Fig. 2. 

  Prescription volume (in millions) from March 2001 through March 2005.

3.2. Predictors supporting observation 

Table 2 presents the estimates for each of the predictor values considered in this study. One predictor that may account for the decreasing pharmacist demand yet increasing prescription volume (Figure 2) was the ratio of pharmacists to technicians. These ratios, which are a function of state laws, decreased by 24% (from 1.15 pharmacists for each technician in 1999 to 0.87 in 2004). This decrease appears to be driven by a 29% increase in number of technicians relative to a smaller increase (9%) in active pharmacists.

Table 2. Longitudinal (1999-2005) estimates of pharmacy demand predictors

The number of new graduates relative to the estimated number of pharmacists switching careers/retiring increased from 2,094 additional graduates in 1999 to 3,058 in 2005. Similarly, pharmacists' median salary increased approximately by 33%, whereas technicians' salary increased by only 23%. Another contributor to the decreasing demand for pharmacists could be the increase in use of mail-order pharmacies, which increased from 4.9% in 1999 to 6.1% in 2004.

3.3. Predictors opposing 

In contrast, other predictors of pharmacist demand indicated that pharmacist demand should be increasing. For example, third-party payers for prescription drug coverage have been estimated to increase the amount of time by 20% for each prescription filled.¹⁷ During the study time period, the use of third-party payers increased approximately by 10%.

Another predictor, the number of registered pharmacies continued to increase between 1999 and 2004. In addition, historical data have indicated that the percentage of pharmacists working within a retail setting has slightly decreased from 60% of all registered pharmacists in 1999 to 55% in 2004. Finally, the amount of money spent on DTC advertising of pharmaceutical advertising increased to 144% between 1999 and 2004.

3.4. Multivariate model result 

The development of the multivariate model failed to result in any of the expected predictor variables being statistically significant after controlling for trend, seasonality, and cycles. Therefore, the results of the multivariate model produced results that were very similar to those developed from the univariate model. As a result, no multivariate forecast models were found to significantly add to the prediction of future pharmacy workforce demand. The lack of robust multivariate models is likely a function of the lack of sensitivity of the predictor variables relative to the ADI estimates.

Back to Article Outline

4. Discussion 

Throughout the historical and forecasted time period, the demand for pharmacists always exceeded supply. However, the demand-to-supply ratio (ADI) appears to be stable after trending downwards for several years; a trend that was validated by predictor variables including the pharmacist-to-technician ratio, number of new pharmacists, new graduates relative to those retiring or switching careers, and the increased use of mail-order pharmacies.

The effect of the implementation of dispensing services through MMA Part D on the trend is not clear, but it appears it will have little or no impact in the long-term. The large confidence interval makes an accurate prediction impossible. Currently, with a little more than 3 months' experience with the new Medicare drug benefit and based on mainly anecdotal evidence, it appears that there has been little impact on the demand for prescriptions. However, the new drug benefit has had a short-term impact on the demand for pharmacists' time; mainly due to a multitude of technical difficulties experienced during the implementation phase of the new benefit. With time, these technical difficulties are being solved and the demand for pharmacists' time should return to previous levels.

When just prescription volume is considered, it does not appear that the implementation of a Medicare prescription drug benefit program will have much impact on future pharmacist demand. However, it must be emphasized that pharmacists are involved in more than just dispensing, and other variables related to those services must be considered. Without considering other services provided by pharmacists and the confounding effects of factors such as pharmacist-to-technician ratios, and the expanding roles and workplace environments where pharmacists are now used; these estimates could potentially lack the robustness necessary to make sound policy decisions related to the pharmacist workforce.

The studies published as part of the Pharmacy Manpower Project have described historical trends using longitudinal methods. Knapp and Livesey³ reported that between 1999 and 2001 the ADI indicated that pharmacist positions were “somewhat difficult” to “difficult to fill,” with regional variation. A more recent publication by Knapp et al⁴ indicated that the demand continues to outpace the supply of pharmacists, but the demand has been moving toward an adequate supply. Finally, because of the potential implications for the profession of pharmacy, it is imperative that the various professional organizations continuously monitor the supply and demand for pharmacists and undertake efforts needed to improve future forecasting efforts.

The implementation of the MMA Part D program is just 1 example of the kind of event that impacts the pharmacy demand-to-supply balance. Because there will continue to be new demands for pharmacists and changes in the supply in the future, new ways to look at workforce demand and supply will be needed. The usefulness of forecasting methods for looking at future pharmacist demand and supply was shown by this study; however, the usefulness of forecasting will be limited until monthly data for key variables are collected.

4.1. Limitations 

Due to limitations of this study, which were beyond the researchers' influence, readers should exercise caution when interpreting the results. A primary limitation is the crudeness of the data available for the predictor variables; these rough measures did not allow for the appropriate development and precision of multivariate forecasting models. The data used in this study were from a multitude of sources and collected by a variety of methods.

All of the quantitative variables thought to influence pharmacist demand obtained for the purposes of this study were measured on a yearly basis. To leverage the sensitivity of the historical ADI estimates, these predictor variables were modified to reflect monthly estimates. This resulted in more limited multivariate models. The additional variables thought to influence the demand for pharmacists were not directly considered in the models because these data were not readily accessible at an appropriate national level or on a longitudinal basis.

Not all potential variables were included in the models and this may have affected results. Such variables were access to pharmacy related technology; complexity of patient's medication therapy; expanded pharmacy store hours; ratio of male to female registered pharmacists; number of part-time pharmacists; expanded role of pharmacy technicians; the expanded role of pharmacists; and expanded employment opportunities for pharmacists in the health care industry, outside of pharmacists' traditional roles.

Finally, as Knapp et al have already stated, difficulties encountered in forecasting are due in part to the use of historical data trends that are based on health care and academic environments that no longer exist.⁴

4.2. Better data are needed 

The lack of available data to better estimate future pharmacy workforce demand is troubling. Data must be developed if the pharmacy profession is expected to confront the uncertainties of a future that could be significantly altered by the implementation of the MMA. The potential increase in demand for prescriptions by Medicare beneficiaries is only one aspect of changes brought about by the MMA. The potential increased demand for drug therapy management provided by pharmacists could also have a significant impact on demand for pharmacists in the retail setting. Therefore, great care is needed in the initial time period after the implementation of Medicare Part D to measure its effect on demand for prescriptions and pharmacists' services so that the potential effect on future demand for pharmacists may be estimated in as timely manner as possible.

The pharmacy profession must not be slow in reacting to future changes in demand. Therefore, professional pharmacy organizations must be proactive in supporting the initiatives necessary to gather and analyze the data required. If the shortage of pharmacists becomes too acute, the health care marketplace will find ways to resolve the shortage that may not be palatable to the profession (eg, increased automation, reliance on technicians, and expanded roles of allied health professionals to include dispensing prescriptions). Equally disconcerting is the potential for a surplus of pharmacists. New pharmacy programs are being developed throughout the country based on the assumption that the shortage will continue.

Back to Article Outline

5. Conclusions 

Based on the results of this study, it appears that the implementation of MMA Part D should have minimal influence on the overall demand for pharmacists. The forecast models indicate that future demand pharmacists will continue to decrease relative to the supply, but caution is recommended when interpreting these results due to the crudeness of the data. A more formal system for capturing monthly data on variables that may more accurately predict the impact on the demand for pharmacists is necessary to allow for more precise forecasting estimates.

Back to Article Outline

References 

1. Gershon SK, Cultice JM, Knapp KK. How many pharmacists are in our future? Bureau of health professions projects supply to 2020. J Am Pharm Assoc (Wash). 2000;40:757–764
   • View In Article
   • MEDLINE
2. Knapp DA. Professionally determined needs for pharmacy services in 2020. Am J Pharm Educ. 2002;66:421–429
   • View In Article
3. Knapp KK, Livesey JC. The aggregate demand index: measuring the balance between pharmacist supply and demand, 1999-2001. J Am Pharm Assoc (Wash). 2002;42(5):391–398
   • View In Article
   • MEDLINE
   • CrossRef
4. Knapp KK, Quist RM, Walton SM, Miller LM. Update on the pharmacist shortage: national and state data through 2003. Am J Health Syst Pharm. 2005;62:492–499
   • View In Article
   • MEDLINE
5. The Pharmacist Workforce: A Study of the Supply and Demand for Pharmacists. Health Resources and Services Administration, Rockville, MD: U.S. Department of Health and Human Services; 2000.
   • View In Article
6. Diebold FX. Elements of Forecasting. 2nd ed.. Cincinnati, OH: South-Western Thomson Learning; 2001;
   • View In Article
7. National Association of Chain Drug Stores Foundation, Chain Pharmacy Industry Profile, (multiple years) 1999-2005.
   • View In Article
8. Bureau of Labor and Statistics, 1999-2004 Occupational Employment and Wages. Available at: http://www.bls.gov/oes/current/oes291051.htm. Accessed 7/25/05.
   • View In Article
9. Bureau of Labor and Statistics, 1999-2004 National Industry-Specific Occupational Employment and Wage Estimates. Available at: http://www.bls.gov/oes/current/naics5_621390.htm. Accessed 7/25/05.
   • View In Article
10. American Association of Colleges of Pharmacy. Profiles of Pharmacy Students, 2004 AACP Institutional Research Report Series. Alexandria, VA: AACP. 2005: Tables 5, 6, 8, 9, 12, 47, 48,and 56.
    • View In Article
11. American Association of Colleges of Pharmacy. Academic Pharmacy's Vital Statistics. Updated March 2005. Available at: http://www.aacp.org/Docs/MainNavigation/InstitutionalData/6676_2005–03.pdf. Accessed 10/3/05.
    • View In Article
12. American Council for Pharmacy Education. Accredited Professional Programs of Colleges and Schools of Pharmacy. Available at: http://www.acpe-accredit.org/students/programs.asp. Accessed 10/3/05.
    • View In Article
13. IMS. Total US spending by type, 2003 Available at: http://www.imshealth.com/ims/portal/front/articleC/0,2777,6599_44304752_44889690,00.html. Accessed 7/28/05.
    • View In Article
14. National Association of Boards of Pharmacy. Survey of Pharmacy Law, (multiple years) 1999-2004.
    • View In Article
15. Walton SM, Cooksey JA, Knapp KK, Quist RM, Miller LM. Analysis of pharmacist and pharmacist-extender workforce in 1998-2000: assessing predictors and differences across states. J Am Pharm Assoc (Wash DC). 2004;44:673–683
    • View In Article
16. Heffler S, Smith S, Keehan S, Borger C, Clements MK, Truffler C. U. S. Health Spending Projections for 2004-2014. Health Affairs. Web exclusive: February 23, 2005. DOI 10.1377/hltaff.W5.74; W5-74 – W5-85. Copyright 2005 by Project HOPE.
    • View In Article
17. Landis NT. National Association of Chain Drug Stores. Non-patient-care activities dilute pharmacists' time, NACDS study shows. Am J Health Syst Pharm. 2000;57:202.
    • View In Article