Accuracy and Validity of Reported Opioid Prescription Days’ Supply

Abstract

Purpose:

The primary objective of this study was to estimate the percentage of opioid analgesic (OA) prescriptions dispensed by Kentucky independent pharmacies with correctly entered days’ supply in the state prescription drug monitoring program (PDMP) system in 2019.

Methods:

Using a two-stage cluster design, pharmacies were sampled with probabilities proportional to the volume of dispensed OAs; 100 random OA prescriptions were sampled from PDMP records submitted by each pharmacy. Following recruitment, demographic information and hard-copy prescription data for sampled records were abstracted on-site. Days’ supply was independently calculated by two pharmacists using a standard formula with disagreements adjudicated blindly by a third pharmacist. Adjudicated days’ supply was compared with that submitted to the PDMP and classified as accurate/inaccurate. Descriptive statistics were used to characterize the sample and a multivariable logistic regression model was used to assess the relationship between accuracy and prescription/practice-related factors.

Results:

A total of 1281 OA prescriptions were reviewed at 13 participating pharmacies. Accuracy of reported OA days’ supply was 89.85%, (95% CI: 86.90, 92.80). Factors associated with accuracy were presence of special instructions from the prescriber [OR 3.13 (95% CI: 1.43, 6.82)], presence of ‘as-needed’ directions [OR 0.29 (95% CI: 0.18, 0.47)], and billing to a third-party payer [OR 1.43 (95% CI: 1.01, 2.02)].

Conclusions:

Accuracy of OA days’ supply reported to the state PDMP was found to be moderately high. Certain prescription-related factors influence accuracy and should be accounted for in future studies. Patterns, including opioid ‘split-billing’ were identified and may impact validity of PDMP and administrative claims studies.

Plain Language Summary

Prescription days’ supply is most often calculated by the pharmacist during the prescription filling process and, in the case of controlled substances such as opioids, is submitted to the state prescription drug monitoring program (PDMP) for documentation and monitoring. Our study collected information directly from 1281 hard copies of prescriptions for opioid analgesics (OA) in 13 independently owned pharmacies in Kentucky. We used this information to calculate days’ supply using a standard formula and compared that to the days’ supply submitted by the pharmacy to the PDMP. We found that days’ supply was 89.85% accurate. The odds of being accurate were much higher if the prescription contained special instructions from the prescriber (e.g., “must last 30 days”) or if the prescription was billed to a third-party payer (e.g., prescription insurance) rather than cash. On the other hand, the odds of being accurate were much lower if the prescription contained instructions to take ‘as needed.’ Future research that uses OA days’ supply should consider the differences in accuracy across these factors.

INTRODUCTION

Concerns about prescribed and unregulated opioid use problems have persisted for decades. As recently as 2019, 70% of drug overdose deaths in the U.S. involved an opioid.¹ While the use of heroin and unregulated synthetic opioids have shown that not all opioid use problems originate from prescriptions, they continue to contribute to opioid overdoses.

In the outpatient setting, days’ supply refers to the number of days for which a prescription is intended to be used by the patient. Prescription days’ supply is infrequently specified by the prescriber. Instead, it is usually calculated by the dispensing pharmacist based on prescribed quantity and directions prior to submission to third-party payers and/or state prescription drug monitoring programs (PDMPs). This calculated days’ supply is commonly used in pharmacoepidemiologic opioid-related research as an outcome^2–5 or exposure.^6–8 For opioid analgesics (OA), days’ supply calculations also have implications on patient care because of limits on the length of initial prescriptions enforced through legislation^2,9–11 and third-party payers.¹² Some increases in opioid prescription volume have been driven by increases in long-term opioid use.^13,14 Additionally, the duration of opioid exposure at the time of the initial prescription is strongly associated with subsequent long-term opioid use.^15–17

Data on the accuracy of days’ supply calculations are lacking. Results from a survey of Kentucky pharmacists showed variation in calculation of days’ supply when presented with two hypothetical OA prescriptions.²⁰ While previous studies have assessed the accuracy of days’ supply regarding osteoporosis medications¹⁸ and inhaled corticosteroids,¹⁹ no studies have assessed the accuracy of reported OA days’ supply.

The primary objective of this study was to estimate the percentage of OA prescriptions dispensed by independent pharmacies in Kentucky and reported to the state PDMP where days’ supply was entered accurately. The secondary objective was to determine pharmacy and prescription factors related to OA days’ supply accuracy.

METHODS

Sampling

A total of 460 pharmacies were identified as independently owned via records provided by the Kentucky Board of Pharmacy. All OA prescriptions filled by independently owned pharmacies were identified within the state PDMP via U.S. Drug Enforcement Administration pharmacy license numbers.

A probability sample of prescription records within independent pharmacies was selected using a two-stage cluster sampling design. At the first-stage, 25 independent pharmacies (clusters) were sampled with probability proportional to the size (PPS), i.e., the volume of OA prescriptions, excluding buprenorphine products, dispensed in 2019 by each pharmacy. At the second stage, a simple random sample of 100 prescription records for OAs within each sampled pharmacy was taken. Non-patient-identifying data elements (e.g., prescription ID, date filled, etc.) were extracted for each prescription from the state PDMP.

Recruitment and Data Collection

Sampled pharmacies were contacted via telephone and/or email to request participation using a recruitment script including elements of informed consent. Pharmacies were offered an incentive for participation of U.S. $500. Thirteen pharmacies agreed to participate for a response proportion of 52%. These pharmacies provided appropriate geographical coverage and variability of prescription volume as per the sampling plan. Participating pharmacies were provided prescription numbers for sampled prescriptions via encrypted, password-protected email so that pharmacy personnel could prepare for the site visit.

Prescription records were reviewed during site visits from July 27, 2021 to August 27, 2021. Physical copies of archived prescriptions were obtained and data elements extracted into REDCap, a secure web-based software platform (licensed from Vanderbilt University by the University of Kentucky) designed to support data capture for research studies.^21,22 The data collection form (Appendix 1) included pre-populated data elements extracted from the state PDMP (e.g., prescription ID, date filled, etc.) to ensure the correct prescription had been retrieved as well as fields to collect additional prescription information (e.g., modality, frequency of use, special instructions, etc.) not available in the PDMP. Additionally, during site visits, pharmacy personnel were surveyed to obtain pharmacy-level characteristics with responses recorded in a separate REDCap form (Appendix 2).

Definitions

Payer source (collected as “combined third party payers”, “private pay (cash)”, or “340b only”) was dichotomized for analysis into “combined third party payers” and “private pay (cash)” as 340b indicates a discounted rate on the prescription purchase and not a filling claim submission. Thus, those billed “340b only” were considered cash payments for analysis. Estimated daily prescription volume of the pharmacy and pharmacist-to-technician ratio were categorized for analysis based on balanced frequencies in each categorized bin.

Calculations

Days’ supply was independently calculated for each prescription by two study pharmacists using a standard formula (Supplemental Figure 1). Days’ supply values were assigned by dividing the number of dosage units dispensed by the frequency of use as directed in the prescription signatura (Sig).²³ For non-whole numbers, days’ supply was rounded up to the full day. For medications prescribed pro re nata (PRN), to be taken ‘as needed’, days’ supply was calculated based on the maximum dosage units prescribed per day, considering the Sig and additional instructions as applicable. To ensure consistency in making decisions, prior to the calculation process, the reviewers completed 10 cases, compared results, and discussed any discrepancies, calibrating their approaches. The percent agreement between study abstractors was calculated. In cases of disagreement, a senior pharmacist adjudicated the calculations. The adjudicated days’ supply values were considered the standard and compared to what had been assigned at the time of dispensing, as recorded in the PDMP. A sampled prescription was categorized as accurate if the prescription days’ supply entered in the PDMP matched the (adjudicated) standard value. If the days’ supply values did not match, the sampled prescription was coded as inaccurate.

Data Analysis

Statistical software used for analyses were RStudio²⁴ (RStudio PBC, Boston, MA, US) and SAS²⁵ (SAS Institute Incorporated, Cary, NC). Unweighted descriptive statistics (mean ± SD, frequency counts and percentages) were used to characterize the sampled pharmacies and PDMP records. To account for the complex sampling design, all further analyses were completed using appropriate sampling weights and SAS²⁵ procedures (SURVEYFREQ and SURVEYLOGISTIC), identifying pharmacies as the primary sampling units (clusters). The weighted proportion of prescriptions with accurate days’ supply was estimated and reported along with its 95% confidence interval (95% CI). A sensitivity analysis was performed to analyze the influence of rounding uneven days’ supply down as opposed to up. The Wald Chi-Square test for independence of row and column variables in a two-way table for complex sampling design was used to assess for significant differences in accuracy among dosage forms and other prescription characteristics. Additionally, a multivariable logistic regression model was used to identify any statistically significant associations between prescription and pharmacy-level characteristics and accuracy of days’ supply reported to the state PDMP, accounting for the complex sampling design. Opioid type (long-acting vs. short-acting), presence of special instructions from the prescriber to the pharmacist, payer source, presence of ‘as-needed’ instructions, federal controlled substance schedule (CII-CV) designation, and presence of a days’ supply checkbox in the pharmacy verification system were included as covariates in the full model. Backwards selection was used to identify a parsimonious final model. Opioid type, schedule, and presence of a days’ supply checkbox were excluded from the final model due to lack of significance. Tests were significant if p-value<0.05. The predictive power of the model measured by the c-statistic was 0.67. Weighted estimated adjusted odds ratios (aODs) and 95% confidence intervals are reported.

RESULTS

A total of 1300 prescriptions were sampled for review at the 13 participating pharmacies. Of these sampled prescriptions, 19 had no complete on-site records available for data collection or yielded incalculable days’ supply due to lack of instructions in the Sig. Reviewers agreed in their calculations for 1243 of 1281 reviewed prescriptions (97%). The remaining 38 records were adjudicated blindly by a third pharmacist. Pharmacy characteristics are shown in Table 1.

Table 1.

Characteristics of Sampled Independently Owned Pharmacies in Kentucky (N = 13)

Demographic	Value
Pharmacists Employed (FTE), mean ± SD	2.65 ± 1.20
Technicians Employed (FTE), mean ± SD	7.65 ± 5.84
Average Prescriptions Filled per Day, mean ± SD	361.15 ± 94.98
Acceptance of Discount Cards, n (%)
Yes	2 (15.4)
No	11 (84.6)
Use of BIN for Cash Prescriptions, n (%)
Yes	7 (53.8)
No	6 (46.2)
Computer System Auto-Calculates Days’ Supply, n (%)
Yes	13 (100.0)
No	0 (0.0)
Verification Checkbox Specific to Days’ Supply, n (%)
Yes	3 (23.1)
No	10 (76.9)

FTE = full-time equivalents

BIN = bank identification number

Of the 1281 reviewed prescriptions, 311 (24.3%) were transmitted electronically to the pharmacy compared to 940 (73.4%) received as written prescriptions. In the PDMP, 1165 prescriptions were reported as third party pay (e.g., insurer, discount program) and 116 were reported as private pay (cash) prescriptions. Of these 1165 prescriptions, review of pharmacy records showed that 49 were private pay prescriptions that were submitted to a private bank identification number (BIN), making the true total of private pay prescriptions 165 rather than 116 (42.2% increase). Independent pharmacies often use private pay BINs to offer their patients a discounted price; while a BIN is used on these prescriptions, they are not submitted to a third party for payment. Data collected regarding prescription Sig revealed 164 prescriptions (12.8%) including a dosage range and 847 (66.1%) designated ‘as-needed’. In addition to the Sig, 320 prescriptions (25.0%) included special instructions from the prescriber to assist pharmacists in understanding prescribing intent. Common special instructions included specified ‘to be filled on’ dates and directions regarding early fills. Quantity prescribed differed from the quantity dispensed (e.g., prescription partially filled) for 35 (2.7%) of prescriptions.

The estimated proportion of 2019 OA prescriptions (excluding buprenorphine products) dispensed by independent pharmacies in Kentucky that had accurate days’ supply entered and reported to the state PDMP was 89.9% (95% CI: 86.9, 92.8) (Table 2). An estimated 3.98% (95% CI: 3.14, 4.82) of the days’ supply entered in the PDMP for OA prescriptions dispensed by independent pharmacies in Kentucky exceeded the accurate days’ supply, including an estimated 1.48% exceeding the accurate days’ supply by only 1 day. An estimated 6.17% of the OA prescriptions in the PDMP had days’ supply lower than the accurate days’ supply, including 3.57% with only 1-day lower days’ supply. The mean weighted difference between days’ supply reported to the PDMP and adjudicated days’ supply was −0.04 days (95% CI: −0.18, 0.09; paired t-test p-value=0.49).

Table 2.

Accuracy of Reported Opioid Analgesic Days’ Supply in State PDMP by Prescription and Pharmacy-related Factors, 2019 (N=1281)

Variable	Unweighted Prescriptions, n (%)	Estimated Percentage of Prescriptions with Accurate Days’ Supply in State PDMP (Weighted Results)
		Accuracy, %	95% CI Lower, %	95% CI Upper, %	p-value1
Overall Accuracy	1281 (100.0)	89.9	86.9	92.8
Opioid Type					0.038*
Long Acting	75 (5.9)	96.0	91.7	100.0
Short Acting	1206 (94.1)	89.5	86.3	92.6
Special Instructions for Pharmacy					0.009*
Yes	320 (25.0)	95.9	93.8	98.1
No	961 (75.0)	87.8	83.7	91.9
Payer Source					0.017*
Private Pay (Cash)	165 (12.9)	86.1	81.8	90.3
Combined Third Party Payers	1116 (87.1)	90.4	87.3	93.5
Written to Take ‘As-Needed’2					<0.001*
Yes	847 (66.1)	86.8	82.7	90.9
No	434 (33.9)	95.9	94.2	97.6
Prescription Modality					0.522
Electronic	311 (24.3)	90.0	85.8	94.3
Facsimile	3 (0.2)	100.0	100.0	100.0
Telephone	24 (1.9)	91.7	78.4	105.0
Transfer	1 (0.1)	100.0	100.0	100.0
Written	940 (73.4)	89.7	85.5	93.8
Unavailable	2 (0.2)
Dosage Form					0.406
Tablet or Capsule	1250 (97.6)	90.0	86.9	93.1
Solution or Suspension	9 (0.7)	66.7	29.1	104.0
Patch	20 (1.6)	95.0	84.2	106.0
Other	2 (0.2)	50.0	50.0	50.0
Prescriber-specified Days’ Supply					0.535
Yes	329 (25.7)	91.5	84.6	98.4
No	952 (74.3)	89.3	86.1	92.5
Federal Drug Schedule					0.155
C-II	1068 (83.4)	90.1	87.1	93.0
C-III	32 (2.5)	75.0	62.5	87.5
C-IV	181 (14.1)	91.2	84.5	97.8
Days’ Supply Checkbox in Pharmacy System3					0.937
Yes	297 (23.2)	89.6	79.7	99.4
No	984 (76.8)	89.9	87.5	92.4
Pharmacist-to-Technician Ratio of Pharmacy					0.277
<0.25	488 (38.1)	91.4	88.5	94.3
0.25–0.49	593 (46.3)	91.4	88.4	94.4
≥0.50	200 (15.6)	81.5	77.5	85.5
Daily Prescription Volume of Pharmacy					0.125
Less than 300	299 (23.3)	92.6	91.4	93.9
300 or greater	982 (76.7)	89.0	85.4	92.6

^*Statistically significant at α=0.05

¹Wald Chi-Square test for independence of row and column variables in a two-way table for complex sampling design

²In ‘as-needed’ treatment, patients are instructed to take only as frequently as their pain dictates.

³Presence or absence of a checkbox in the pharmacy’s prescription processing software specifically requiring the pharmacist to verify days’ supply during prescription verification

Results of the multivariable logistic regression modeling are shown in Table 3. Presence of special instructions from the prescriber for the pharmacist was associated with 3.1 (95% CI: 1.4, 6.8) times greater adjusted odds of accurately calculated days’ supply, adjusting for other significant covariates. Additionally, prescriptions billed to third parties were found to have 1.4 (95% CI: 1.0, 2.0) times greater adjusted odds of accurately calculated days’ supply. Presence of instructions for the patient to take ‘as needed’ was associated with 71% lower adjusted odds of accurately calculated days’ supply.

Table 3.

Factors Associated with Increased Accuracy of Reported Opioid Analgesic Days’ Supply in State PDMP, 2019

Variable	Adjusted Odds Ratio (95% CI)2
Special Instructions for Pharmacy
No	Reference
Yes	3.13 (1.43 – 6.82)*
Payer Source
Private Pay (Cash)	Reference
Combined Third Party Payers	1.43 (1.01 – 2.02)*
Written to Take ‘As-Needed’1
No	Reference
Yes	0.29 (0.18 – 0.47)*

^*Statistically significant at α=0.05

¹In ‘as-needed’ treatment, patients are instructed to take only as frequently as their pain dictates.

²Results from a multivariable logistic regression accounting for the complex sampling design

There were 130 prescriptions in the sample with inaccurately entered days’ supply in the state PDMP. A records review was performed to identify sources of inaccurate days’ supply calculation. The estimated proportion of inaccurate OA prescriptions with underreported days’ supply (i.e., days’ supply reported to PDMP was less than calculated) was 60.0% (95% CI: 49.9, 70.1). Manual record review of 28 inaccurately reported prescriptions with prescriber-specified days’ supply revealed a pattern of pharmacists defaulting to the prescription processing software’s calculated days’ supply even when the prescriber had specified the prescription was to last longer. Additionally, a small number (n=3; 10.7%) of these prescriptions were billed as a 7 days’ supply, a limit often imposed by third parties, despite differing specifications from the prescriber.

Data abstraction revealed several patterns in days’ supply reporting. The primary observation made by abstractor was the presence of ‘split-billing’ in the sample. We define ‘split-billing’ as when a filling pharmacy, to circumvent initial opioid fill days’ supply limits, initially fills the covered days’ supply (typically 7 days) and submits to insurance. On the same day as this insurance fill, the pharmacy then fills the remaining supply as private pay (cash). Both fills are then reported to the state PDMP as two separate fills. Of 1281 reviewed prescriptions, 16 (1.25%) were noted by the abstractor to be ‘split-billed’.

Finally, regarding sensitivity analysis, the rounding down of calculated days’ supply yielded an estimated accuracy of 87.1% (95% CI: 83.0, 91.2).

DISCUSSION

This study shows that, overall, the accuracy of OA days’ supply reported to the state PDMP is high. However, accuracy is considerably lower for prescriptions written to be taken ‘as-needed’, while prescriptions with special instructions from the prescriber and prescriptions billed to a third party were associated with greater odds of accuracy.

Accuracy of OA days’ supply plays a crucial role in pharmacoepidemiologic studies, extending beyond its use as a direct outcome or exposure as it also plays a critical role in the calculation of morphine milligram equivalents (MME), which are used to standardized differences between opioid molecule potency.²⁶ Previous research seeking to illustrate the variation in MME introduced by methodological differences in calculation procedures found that, across the 4 most common differing procedures, variations in calculated MME were three-fold.²⁷ Notably, 2 of the 4 procedures identified for MME calculation utilize days’ supply in the denominator, illustrating the importance that accuracy of OA days’ supply may have on MME calculation, with corresponding impact on interpretation of opioid-related studies.

Implications of the accuracy of OA days’ supply reporting also extend to clinical practice. In 2016, population-level mortality data led to changes in U.S. Centers for Disease Control and Prevention (CDC) guidelines, including a strong caution against chronic noncancer pain management of opioid doses above 90 daily MME.²⁸ The same guidelines also cited a lack of evidence regarding superiority of dosing intervals, including scheduled dosing as opposed to ‘as-needed’ dosing. Inaccuracies in OA days’ supply among those written to take ‘as-needed’ could impact the interpretation of findings to clinical practice and possibly attenuate mortality and other clinically relevant outcomes.

An important consideration for future research is days’ supply accuracy by payer source. It is unsurprising that accuracy of days’ supply is higher for OA prescriptions billed to third-party payers as incorrect days’ supply can cause claim rejections and might raise audit red flags.²³ Results of this study suggest that accuracy of reported OA days’ supply in studies utilizing insurance claims data may be higher (when captured) than that of studies utilizing PDMP data; the latter captures private pay prescription fills in addition to those submitted to third parties. These nuances are particularly important because private-pay opioid prescriptions have been found to be associated with 30% higher mean MME values²⁹ and a 23% lower odds of naloxone co-prescribing compared to commercially insured encounters.³⁰ Additionally, patients exhibiting ‘opioid shopping behavior’ were found to be more likely to pay in cash (private pay).³¹ Inaccuracies in days’ supply of private pay OA prescriptions thus introduces a crucial hurdle to overcome in future studies.

We also document differential misclassification among private pay prescriptions. As mentioned previously, independent pharmacies often use private pay BINs to offer their patients a discounted price; while a BIN is used on these prescriptions, they are not submitted to a third-party for payment. As such, these private pay prescriptions are classified as third-party in the PDMP and not as cash. Accounting for this misclassification yielded a 42.2% increase in the number of private pay OA prescriptions. This calls into question the validity of using payer source in research and clinical risk prediction algorithms implemented with PDMP data. Additionally, as only 2 of 13 sampled independent pharmacies accepted discount cards, this study did not assess the accuracy of OA prescriptions billed to discount cards/savings program BINs, which may further add to the ambiguity surrounding payer source. These prescriptions, like those submitted to a cash BIN, are not subject to third party audits and rejections regarding incorrect days’ supply, and thus may be subject to the same decrease in reported days’ supply accuracy. Without a means of differentiating true private pay prescriptions from third party payers in large databases such as state PDMPs, interpretation of payer source in opioid-related studies may be limited.

Inaccuracies among prescriptions with prescriber-specified days’ supply were most commonly due to pharmacists submitting a lower days’ supply value calculated by the prescription processing software, rather than that specified by the prescriber. An example scenario is a prescription written for 75 tablets with directions to ‘take 1 tablet 3 times daily as-needed’ and a prescriber-specified days’ supply of 30 days. In this scenario, the prescriber intends the ‘as-needed’ prescription to last 30 days, although if the patient took the maximum amount each day, the prescription would only last 25 days. The prescription processing software, once the prescription sig is entered, automatically calculates the days’ supply as 25 days. When not manually changed to 30 days by the pharmacist, this results in an underreported days’ supply.

The estimated proportion of inaccurate OA prescriptions with underreported days’ supply was 60.0% (95% CI: 49.9, 70.1). This inaccuracy in reporting leads to a situation where proportion of days covered (PDC) may be artificially lower and MME calculation may be artificially higher, given that the same amount of opioid is intended by the prescriber to last a longer time than reported. This has implications for gap allowances used in pharmacoepidemiology studies assessing continuous dosing intervals.

Prior to this study, the practice of prescription opioid ‘split-billing’ had not been described. Depending on the methodology used to classify opioids, ‘split-billing’ may lead to a chronic opioid fill having the appearance of an acute fill given its shortened days’ supply. Additionally, the portion of the fill that is billed as private pay is censored from administrative claims data, further clouding the true clinical picture of the patient. Further large-scale studies are necessary to explore this practice in other populations. While Kentucky legislation allows for the practice of ‘split-billing’, legislation in other states or territories may prevent controlled substances from being filled twice on the same date, impacting generalizability.

Our study has limitations. As our focus was independently owned pharmacies, results may not be generalizable to other pharmacy types (e.g., chain pharmacies, mail-order pharmacies) that may employ different workflows and prescription processing software. Additionally, response bias is a possibility given the nature of pharmacy recruitment during the SARS-CoV-2 pandemic. Due to the high response proportion and reported reasons for refusal to participate (e.g., shortage of staff, priorities, no prior engagement in research), we do not believe refusals lead to systematic bias.

Overall accuracy of OA days’ supply reporting in the state PDMP was found to be high. Factors such as special instructions from the prescriber, ‘as-needed’ directions, and payer source were, however, found to influence accuracy of reported OA days’ supply. In future studies where OA days’ supply is utilized, factors such as these should be accounted for in sensitivity analyses and result interpretation. Additionally, large-scale studies are necessary to further explore the prevalence of ‘spit-billing’ so that studies exploring OA days’ supply can better discern between acute fills and ‘split-billed’ prescriptions.

Key Points:

• Prescription days’ supply is infrequently specified by the prescriber, instead calculated by the dispensing pharmacist

• Overall accuracy of opioid analgesic (OA) days’ supply reporting in the Kentucky state PDMP was found to be high

• Presence of special instructions from the prescriber and third-party prescription payer source were associated with increased accuracy of reported days’ supply while ‘as-needed’ directions were associated with decreased accuracy.

• In future studies where OA days’ supply is utilized, factors such as these should be accounted for in analyses and interpretation of results

• The practice of opioid ‘split-billing’ was revealed in the sample and may obscure prescription details in both PDMPs and administrative claims databases

Appendix 1: Prescription Data Collection Form

Record ID:
Read-Only KASPER Data	Data Entry
Prescription ID:
Sample:
Abbreviated pharmacy name:
Pharmacy DEA ID:
Payer source: Private pay (cash) Other	Payer source from pharmacy: Private pay (cash) Other 340-B (no additional third party)
Date filled:
Drug name:	Is drug name from KASPER same as RX? Yes No
Dosage form from KASPER:	Is dosage form from KASPER same as dosage form from RX? Yes No
Quantity dispensed from KASPER:	Is quantity dispensed from KASPER same as RX? Yes No
	Is quantity PRESCRIBED on RX same as quantity DISPENSED from KASPER? Yes No
	Directions for use from RX:
	Re-enter directions for use from RX:
	Did the prescriber specify a days’ supply? Yes No
	If the prescriber specified a days’ supply, enter it here:
	Do directions for use contain a range? Yes No
	Does the drug contain acetaminophen? Yes No
	Do directions for use include ‘as needed’? Yes No
	RX origin from pharmacy: Electronic Written Phone Facsimile Transfer
	Does the RX have special instructions? Yes No
	Reviewer Notes:
	Days’ Supply Calculation: (to be entered after data abstraction)

Appendix 2: Pharmacy Data Collection Form

1. What is the name of your pharmacy? ______________________

2. What are your pharmacy’s operating hours? (please include all 7 days of the week) ______________________

3. How many pharmacists are employed at your pharmacy? (full-time equivalents) ______________________

4. How many pharmacy technicians are employed at your pharmacy? (full-time equivalents) ______________________

5. On a daily basis, what is the average staffing composition for your pharmacy? (e.g., 1 pharmacist and 3 technicians per day) ______________________

6. On average, how many prescriptions does your pharmacy fill per day? ______________________

7. What software/system do you use in your pharmacy? ______________________

8. Does your pharmacy dispense medications under the 340B program?

   1. Yes
   2. No
9. When processing prescriptions under 340B, do you use a specific BIN to process the prescriptions?

   1. Yes
   2. No
10. Does your pharmacy accept discount cards?

    1. Yes
    2. No
11. When processing prescriptions that are NOT billed to a third party (i.e. when the patient pays cash), do you use a specific BIN to process the prescription?

    1. Yes
    2. No
12. Does your pharmacy software/system automatically calculate the days’ supply based upon the SIG and quantity entered?

    1. Yes
    2. No
13. When verifying a prescription, does your pharmacy software/system require you to SPECIFICALLY check the days’ supply (e.g., a checkbox to confirm days’ supply is accurate)?

    1. Yes
    2. No

14. Reviewer Notes: ______________________