KASPER Evaluation Team*

Transcription

1 Study of Real-Time Prescription Monitoring Program (PMP) Data Collection for the Kentucky All Schedule Prescription Electronic Reporting (KASPER) Program KASPER Evaluation Team* Institute for Pharmaceutical Outcomes and Policy Department of Pharmacy Practice and Science College of Pharmacy University of Kentucky Lexington, Kentucky June 2011 * Patricia R. Freeman, PhD, RPh Karen Blumenschein, PharmD Jeffery Talbert, PhD Stacy Taylor, PharmD, MHA Joseph L. Fink III, BSPharm, JD

2 CONTENTS I. Introduction..3 II. III. IV. Scope of Work..5 State PMP Stakeholder Interview..6 Independent Pharmacist Stakeholder Interview V. Chain Pharmacy Manager Interviews. 11 VI. VII. Survey of Non-Pharmacy Dispensers of Controlled Substances...14 Software Vendor Interviews VIII. Summary and Recommendations IX. Appendices

3 I. Introduction Since the early 2000s, the mandate for and implementation of prescription drug monitoring programs (PMPs) has become a national movement. As of May 16, 2011, 49 states have mandated programs and 35 of these states have implemented operational PMPs. The common goal among these programs is to reduce prescription drug abuse and diversion by discouraging doctor shopping, which occurs when patients see multiple providers and pharmacies with the intent of obtaining controlled substances for misuse and/or diversion. This goal should be accomplished without causing a chilling effect that compromises patient access to controlled substances (CS) required for legitimate medical needs. However, there is wide variation in program design, objectives, and operation across the states, including the mechanism and the frequency with which CS prescription data are reported from dispensers to the PMP. Currently in the majority of states, dispensers report CS prescription data to the PMPs via electronic batch files. Aggregate data are collected on the CS prescriptions dispensed within the dispensing system over a given time period and then a batch file is created from the data and uploaded to the state PMP database. The common procedure for processing and transmitting CS prescription data to PMPs is depicted in Figure 1 below. Figure 1: Controlled Substance Prescription Data Reporting Process 3

4 Kentucky s PMP, the Kentucky All Schedule Prescription Electronic Reporting Program (KASPER), has contracted with RelayHealth as the vendor for all CS data collection. Currently, in addition to electronic batch file transmission as described above, RelayHealth offers an option to automatically extract a pharmacy s CS reporting information directly from the adjudication transactions sent to the RelayHealth switching network which eliminates the need for the pharmacy to submit batch files 1. The majority of states (32) mandate data reporting every 7 14 days (Figure 2), including Kentucky which currently mandates reporting of CS data to the PMP every seven days. Three states - Minnesota, North Dakota and Oklahoma - currently require dispensers of CS to report data to the PMP daily. To date, only one state - Oklahoma - has mandated real-time reporting to the PMP effective January 1, Currently, there is no standard definition of real-time. The Oklahoma legislature has defined real-time as at the point-of-sale, i.e., at the point the CS prescription is purchased and placed in the hands of the patient consumer 2. Another view of real-time is at the point of CS prescription filling and adjudication, i.e., at the time the CS prescription is entered into the pharmacy dispensing systems and processed for third party payment through the switch as described in the American Society for Automation in Pharmacy (ASAP) Standard for Prescription Monitoring Programs % 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Figure 2: Mandated Frequency of Data Reporting to PMPs 4 n=40 1 Relay Health Prescription Monitoring Program Information, accessed May 24, 2011, 2 Oklahoma Bureau of Narcotics and Dangerous Drugs Control Prescription Monitoring Program, accessed May 24, 2011, 3 Implementation Guide ASAP Standard Version 4 Release 1 for Prescription Monitoring Programs, accessed June 11, 2011, 4 Alliance of States with Prescription Monitoring Programs, accessed May 24, 2011, Monthly Biweekly Weekly Daily 4

5 Proponents of real-time CS prescription data reporting argue that healthcare providers need current real-time data in PMP reports so they can make accurate CS prescribing and dispensing treatment decisions. In contrast, opponents of real-time data reporting have suggested that the costs associated with implementing the technological enhancements required for real-time data reporting are too high to outweigh the benefits gained by real-time transmission as compared to more frequent batch transmissions, i.e. daily. Accordingly, this present study sought to identify the barriers to real-time CS prescription data submission to the KASPER program and to gather opinions on the advantages and disadvantages of real-time reporting from KASPER stakeholders. II. Scope of Work The KASPER Evaluation Team was engaged by Cabinet for Health and Family Services (CHFS), Office of the Inspector General, to conduct a study to identify the challenges and costs associated with implementing real time data collection within the KASPER program. To accomplish this, the following study components were outlined in the scope of work to be conducted by the Team: 1) Conduct a survey of state PMPs and identify and interview stakeholders in states identified as moving quickly toward real-time PMP data collection. 2) Develop and provide a continuing education program on PMPs, including approaches to and advantages of real-time reporting in terms of clinical decisionmaking at the point of care. 3) Conduct a stakeholder interview with independent pharmacy providers from across the state to get input on real-time data collection. 4) Conduct stakeholder interviews with representatives from a minimum of five major pharmacy chains servicing Kentucky. 5) Conduct a survey of a minimum of 10 non-pharmacy dispensers of CS, including dispensing physicians, veterinarians and dentists. 6) Conduct a survey or interview with seven major pharmacy software vendors. 7) Identify policy issues related to real-time data collection within the KASPER program and recommendations to address these issues. 5

6 III. State PMP Stakeholder Interviews The KASPER evaluation team participated in a project kick off meeting via teleconference with KASPER and CHFS staff on July 1, Information provided by CHFS staff indicated that Oklahoma is the only state moving toward real-time data collection and provided contact information for the Oklahoma PMP manager. The KASPER evaluation team conducted a phone interview with the Oklahoma PMP manager on September 23, 2010 and posed the following questions: 1) How is OK defining real-time transmission? 2) What is OK s approach to real-time transmission? a. What is the platform for transmission? b. Who is required to transmit real-time (all or just pharmacy dispensers)? 3) What is current timeline for implementation of real-time transmission? 4) What input did OK obtain from relevant groups (if any) before moving toward real time? From whom did OK seek information? How was the information solicited/collected? 5) What obstacles/challenges have been encountered along the way? How did these impact the implementation timeline? 6) How were the obstacles/challenges addressed? 7) What is the best estimate of the cost to pharmacies and the state of OK for moving to real time transmission? Real time transmission in Oklahoma is defined as the point-of-sale i.e., within 5 minutes of the actual time at which the CS prescription is put into the hands of the patient/consumer and not at the point of processing and adjudication. All dispensers, regardless of practice site, will be required to transmit real time to the Oklahoma PMP. To get buy-in for the move to real-time data transmission, a planning committee was created that included representatives from major chain pharmacy and dispensing system software vendors. The planning committee worked with PMP staff to identify an appropriate timeline for implementation that would allow for developing and implementing the needed technological upgrades that would be compatible with realtime transmission at the point of sale. The planning committee s work coincided with the legislative initiative that would statutorily mandate real-time transmission by January Although little resistance was encountered during the legislative process, as the deadline for implementation (January 2012) grew nearer, some chain pharmacies raised issues relative to their ability to meet the statutory requirements. Oklahoma PMP utilizes software developed in-house as the platform for data transmission so there is no vendor and therefore no external costs associated with implementation. A web form has also been created for practitioners who dispense small volumes of CS so that CS prescribing data can be entered manually instead of through direct electronic transmission from dispensing system software. 6

7 IV. Continuing Education Program-Independent Pharmacist Stakeholder Interviews The KASPER Evaluation Team developed and conducted a 1.5 hour continuing education (CE) program entitled The Future of KASPER: Technology Enhancements at the 13 th Annual American Pharmacy Services Corporation (APSC) Meeting for Independent Pharmacy, on Friday, August 6, 2010 from 1:00 2:30 PM. The CE program, presented by Dr. Patricia Freeman of the KASPER evaluation team, was presented in two parts. Part I was designed to provide the participants with an overview of prescription monitoring programs and to discuss the findings of the pharmacist survey conducted as part of the recent KASPER evaluation conducted by the KASPER evaluation team. Part 2 was designed to collect information relative to pharmacists opinions on real-time transmission of CS data to the KASPER program and served as the stakeholder interview component of this study. The stakeholder interview protocol was approved by both CHFS and University of Kentucky (UK) Institutional Review Boards. A full copy of the slides, which were distributed as a handout to audience participants, can be found in Appendix 1. The learning objectives for the program were as follows: Part I: - Describe the current status of prescription monitoring programs in the US - Discuss findings from the recent evaluation of the KASPER program Part II: - Identify the advantages of real-time reporting of CS prescription data - Discuss the technology enhancements needed to accomplish real-time data reporting - Identify obstacles and barriers to the implementation of real-time data transmission In addition to KASPER evaluation team members and APSC staff, 43 participants were in the audience. Following the presentation of Part I, demographic information was collected from the audience using the Turning Point Audience Response System. Twenty-two percent identified themselves as a pharmacy owner, 48% identified themselves as owner and pharmacist, 19% as pharmacist, and 11% as pharmacy technician. The participants had been in practice an average of 27 ( ± 14) years and dispensed an average of 257 ( ± 113) prescriptions per day. Of these 257 prescriptions, the participants estimated they dispensed an average of 55 ( ± 43) CS prescriptions daily. The majority of participants were from the North Central region of Kentucky (48%) while 33% indicated they were from the eastern region, 11% from the western region and 7% from the south-central region as depicted in the graphic depicted in Figure 3 below. 7

8 Figure 3: Geographic Regions of Kentucky When asked, 77% of participants reported having registered for a KASPER account while 19% did not have an account; 4% were unsure. Sixty-seven percent of participants indicated they had personally transmitted CS prescription data to KASPER. The majority of participants (79%) identified third-party insurance as the primary payer for prescriptions in the pharmacy while 21% of participants indicated Medicaid as the primary payer. Rx30 was the most commonly reported dispensing software with 50% of participants indicating they used this system; other systems participants reported using include QS1, HCC, HBS, Computer Rx, Renlar, RxKey and SRS. When asked if their prescription dispensing system was capable of real-time data transmission at the point of controlled substance dispensing, 46% indicated yes, 11% indicated no, while 43% indicated they were not sure or did not know. Seven distinct discussion questions were asked during Part II of the program which generated much discussion. The participants were very engaged in the discussion, sharing ideas and opinions freely. The major themes that emerged from each of the questions are summarized below. How is controlled substance prescription information currently transmitted to KASPER from your pharmacy? Does this transmission require any actions on your part (or by someone else in the pharmacy)? Some participants reported using batch files as the mode of weekly transmission while others expressed using the automatic extraction process previously offered by RelayHealth. The general consensus of those that used batch reporting was that the pharmacist had to remember to build/save/send the data report weekly. In contrast, those that used the automatic extraction feature had to remember to flag all CS prescriptions that are cash transactions rather than third-party transactions so that they are transmitted to RelayHealth for automatic extraction. [RelayHealth instructions for submitting electronic batch files and for automatic CS data extraction accessed from the RelayHealth website on May 16, 2011 are provided in Appendices 2 and 3 for reference]. 8

9 Several participants voiced concern over the fact that automatic extraction of data from claims by RelayHealth occurs at the point of adjudication and not at the point of sale. Thus, claims for prescriptions that are adjudicated but never picked up are still submitted to KASPER. This was specifically identified as a concern with possible real-time transmission of data to KASPER at the point of adjudication according to one participant who said There needs to be a way to pull (the prescription) off the KASPER report if you reverse a cash prescription. If your prescription dispensing system is not capable of real-time data transmission at the point of controlled substance dispensing, what needs to be changed in order to accommodate this? Participants seemed to indicate this was more of an issue with the adjudication switch companies e.g. RelayHealth rather than an individual software issue, although as stated above some participants indicated the software they used was not capable of real-time data transmission. Concern again was voiced regarding transmission at point of dispensing (adjudication) vs. point of sale. Debate ensued over the value of transmission at these two different points in the dispensing process and no consensus was reached by the audience on the benefits/merits of point of sale vs. point of dispensing transmission. Of interest to note is that not all pharmacies currently have point-of-sale systems that are integrated with their dispensing systems. When polled, approximately two-thirds of the participants in the audience indicated they currently have point-of-sale systems. What costs do you incur under the current system that requires data transmission to KASPER every seven days? How would these costs change if KASPER moved to real-time data transmission? Kentucky law mandates that KASPER must not charge a fee to use the system. This mandate is included in all information provided by Kentucky s data collection vendor (RelayHealth) to dispensers of CS. It was clear from the discussion that some pharmacist dispensers believe they are paying fees to the switch for CS data reporting. Individual responses to this question included: For cash prescriptions we have to pay to go through the switch. For real time we would have to pay for each cash prescription we transmit and it is 5-7 cents for each one. We pay 6 cents per transaction and we get billed for reversals. It should be zero cost why would the burden be on us (meaning pharmacists)? KASPER gets a free-ride because we are doing this stuff for other payers. Only one participant offered the comment that Right now it is free to KASPER. 9

10 What do you see as the biggest obstacle to real-time data transmission at the point of controlled substance dispensing? The general consensus of participants in response to this question was the ability to have and interface with the appropriate technology that would allow for realtime data transmission. Not being sure of the technology required to accomplish real-time, one participant indicated that the potential cost associated with implementing the required technology would be an obstacle. An interesting comment relative to real-time transmission is that if it is going to be implemented it should be interactive. The system should not just be designed to gather data with no feedback to the pharmacists. As one participant stated: You can t run a KASPER every time you fill a controlled substance so the system has to be interactive. By interactive, pharmacist participants meant that any potential real-time requirement for CS data transmission to KASPER should provide immediate feedback to pharmacists similar to the drug utilization review (DUR) alerts pharmacists receive when adjudicating Medicaid claims. An additional concern raised in response to this question centered around pharmacists liability. If they do this will that change our liability? Right now KY has no liability will that change? Will we get sued for cutting people off? What do you perceive as the advantages of real-time data transmission at the point of controlled substance dispensing? Responses to this question were varied. Some participants questioned the advantage of real-time, That s what I want to know, what IS the advantage of real- time? I don t see it. Others suggested that the real advantage would be if the system were two-way data were transmitted and information received back from KASPER in real-time like a DUR alert as previously mentioned. What do you perceive as the disadvantages of real-time data transmission at the point of controlled substance dispensing? Participants suggested that the main disadvantage of real-time data transmission at the point of dispensing is related to the technological changes that would be needed. Some people may have to change software. Some will have to pay $10K for a software upgrade. Additionally, comments were made questioning the role of the pharmacist vs. the prescriber in ensuring appropriate CS use. The state should tell prescribers to transmit KASPER data --- remove us (meaning pharmacists) from being the middle man 10

11 Do you have any suggestions regarding how KASPER could obtain controlled substance data from Kentucky pharmacies in real time? The specific responses provided by the participants did not really address the question posed. Participants continued to question the value of the pharmacist as the health care professional policing this issue vs. the prescriber. KASPER should be run before prescribing not dispensing. KASPER should be part of the standard of care for chronic pain patients. V. Chain Pharmacy Manager Interviews Six major chains which own and operate pharmacies in Kentucky were identified for interviews. To gather input from chain pharmacy managers, the Kentucky Retail Federation was approached for assistance in identifying appropriate individuals at the corporate level for interviews. Contact information for chain pharmacy managers was obtained and s requesting participation in the telephone interviews were sent by the KASPER evaluation team. Positive responses, implying consent to participate in the study, were received from four of the six chains identified. Telephone interviews with the chain pharmacy mangers were conducted during November January 2011 by members of the KASPER evaluation team. A series of standard questions, similar to those posed to the independent pharmacists during the APSC stakeholder interview in August, were presented to the participants. The full set of questions is included in Appendix 4. Combined, the four chains operated over 400 pharmacies geographically dispersed across Kentucky which dispensed on average approximately 300 prescriptions per pharmacy daily. As with the independent pharmacist stakeholder interviews, the responses to the questions were documented and reviewed for themes that emerged across all four interview sessions with the chain pharmacy managers. The major themes that emerged from each of the questions are summarized below. How is controlled substance prescription information currently transmitted to KASPER from your pharmacy? Does this transmission require any actions on your part (or by someone else in the pharmacy)? Representatives from all four chain pharmacies indicated that CS prescription information is transmitted via batch file through RelayHealth to the KASPER program and that this process is done behind the scenes with no action required on the part of the pharmacist at the pharmacy level. Currently, Kentucky law mandates transmission of CS data to the KASPER program every seven days, two of the chains reported transmitting data to KASPER on a more frequent basis (twice a week and daily). Is your prescription dispensing system capable of real-time data transmission at the point of controlled substance dispensing? If not, what needs to be changed in order to accommodate this? 11

12 Representatives from all four chains answered this question with a resounding no. When questioned further, all said while the technology is available that would allow for real-time transmission of CS prescription data to KASPER, changes to the dispensing system software would be required. I can answer this for the entire industry NO. Obviously we could, I mean there are some triggers we could add.but that would be a significant development effort and I m not sure what that gets us. Another participant stated We would have to make changes to it. Most of the information we consider real-time today is adjudication. Oklahoma is looking at it right now and we have to make changes. I would question the reason why I don t understand why it would need to be real-time versus a daily batch transmission. Additionally, much discussion on this question surrounded the definition of realtime and the problems that would be encountered during the dispensing process workflow should real-time be defined as the point-of-sale, rather than at the point of prescription processing and adjudication. Clearly if real-time is defined as point-of-sale vs. point-of-adjudication, then costs associated with purchasing and using integrated point-of-sale systems would be incurred by those pharmacies that currently do not utilize them. Concern was voiced regarding transmission of data at the point of processing/adjudication, which then would be available immediately to the provider via KASPER reports and could present misleading information upon which a provider might take unnecessary action. There are a multitude of definitions of dispensing. The step of adjudication can be viewed as real-time but when exactly is dispensing? When you hand it to the customer? Which step in the process is real-time.at point of sale or when it s adjudicated? What costs do you incur under the current system that requires data transmission to KASPER every seven days? How would these costs change if KASPER moved to real-time data transmission? All agreed there would be a cost associated with moving to a real-time or near real-time model of CS prescription data transmission to KASPER. While individual participants from the various chains estimated these costs differently, the general consensus of this group of stakeholders was that the costs would be associated with building the IT solutions to accomplish real-time, once built and deployed, routine maintenance costs would not be that different than they are currently. There s not so much of a cost once the infrastructure is there to support it. The IT resources are a cost not so much to get it deployed but to build it..i would say it would be north of $100,000 to build this. Another participant responded The development work is what would be expensive. I don t think daily would be nearly as difficult as real-time. When we went from weekly to twice a week it was relatively smooth. 12

13 What do you see as the biggest obstacle to real-time data transmission at the point of controlled substance dispensing? In response to this question, a couple of participants first replied: Why is it needed? What is the rationale behind the move to real-time? They went further to express their concerns relative to the role of the pharmacist as a law enforcement official. Is the pharmacist going to get information back in realtime so that they can act as a policeman are they supposed to stop that prescription that they catch from going through? Additionally, one participant questioned why the state was focusing on catching the 1% of the population that attempt to get CS drugs for non-medical needs. What do you perceive as the advantages of real-time data transmission at the point of controlled substance dispensing? None of the four chain participants articulated any advantage to real-time data transmission. I don t really see an advantage to real-time we are health care providers. Nowhere in the definition does it say we are law enforcement. Another participant stated I don t see an advantage the way that it is being proposed. What do you perceive as the disadvantages of real-time data transmission at the point of controlled substance dispensing? Participants suggested cost and the IT resources it would take to develop a system compatible with real-time reporting. Participants continued to question the value of real-time over daily transmission. It would help if we could understand what the benefits of it are. Is someone going to be acting on this real-time? Otherwise what s the difference between real-time and nightly? Real time sounds good but I am worried about the unintended consequences. A legitimate patient could be denied care because of interpretation of what the pharmacist sees in the data. What do you do with someone who has a reaction with one drug and so goes back to the doctor the same day and it looks bad to you on a KASPER report, but it s really a normal situation. Do you have any suggestions regarding how KASPER could obtain controlled substance data from Kentucky pharmacies in real time? The consensus of the participants again was related to understanding why realtime is advantageous over daily reporting. Additionally, participants indicated that if the state were to mandate real-time reporting, it should be done in a way that would not impact workflow. One participant stressed the need for ample implementation time. Once you define real-time, the implementation has got to be a minimum of six months Ultimately, although participants questioned the 13

14 value of real-time, the general consensus was that if it were mandated, it could be done. If the legislature says real-time has to be done then we will comply. VI. Survey of Non-Pharmacy Dispensers of Controlled Substances The purpose of the survey was to gather information from non-pharmacy dispensers of CS, including dispensing physicians, dentists and veterinarians, to identify the effort and costs associated with providing data real-time to the KASPER program. The survey questions were developed with input from KASPER and CHFS staff and copies of the survey instruments are provided in Appendices 5 and 6. The survey protocol was approved by both the CHFS Institutional Review Board (IRB) and the University of Kentucky IRB. A. Survey Methodology The survey methodology followed a slightly modified version of the method described by Dillman 5. The sample of physician dispensers included all physicians reporting CS dispensing to the ekasper program through RelayHealth on July 14, A total of 24 unique physician dispensers were identified in the provided dataset. Due to the small number of individuals in the sample, all identified physician dispensers were mailed a survey packet containing a plain text survey form and a business reply envelope with return postage. One week following the mailing of the survey packet, a postcard was sent that thanked those who had returned the questionnaire and asked those who had not to please do so. Anyone who had not responded after the postcard wave was then sent a second copy of the questionnaire packet. This second mailing occurred two weeks after the first survey was sent. The veterinarian sample included all veterinarians reporting CS dispensing to the ekasper program through Relay-Health on July 13, A total of 47 unique veterinarian dispensers were identified in the provided dataset and were mailed surveys as described above. Although the original scope of work called for the administration of a survey to dispensing dentists, the KASPER program manager indicated that no dispensing dentists were registered with KASPER to report CS dispensing as of July 14, B. Survey Findings 1. Response Rates Seven surveys sent to physician dispensers were returned due to invalid addresses, decreasing the final number of physician dispensers in the sample from 24 to 17. Four responses were received from dispensing physicians for a response rate of approximately 24%. 5 Dillman DA. Mail and telephone surveys: the total design method. New York: John Wiley & Sons,

15 Seven surveys sent to veterinarian dispensers also were returned due to invalid addresses decreasing the final number of veterinarian dispensers in the sample from 47 to 40. Twenty-six responses were received for a response rate of approximately 65%. Overall response rate from the two groups was 53% and is similar to response rates reported for other surveys of health care providers. Complete summary tables of survey responses by group (physician and veterinarians) are included in Appendices Dispensing Physician Responses Four (75%) responding physicians reported dispensing CS prescriptions daily whereas 1 (25%) reported dispensing CS prescriptions one or more times per week. The reported number of CS prescriptions dispensed per month ranged from to greater than 100. All dispensing physicians that responded to the survey reported they frequently (75%) or always (25%) requested a KASPER report of a patient s CS prescription history prior to dispensing. All responding physicians indicated they currently submit CS dispensing information to KASPER on a weekly basis and all report having internet access at the practice site via desk-top and/or laptop computers. Two respondents (50%) indicated they submit data electronically while one respondent (25%) listed completing the approved paper form and submitting via fax to RelayHealth as the method of data reporting. One respondent did not answer this question. When asked about the costs associated with the current method utilized for data reporting to KASPER, none indicated any costs incurred other than personnel time required to complete and submit the data report. When asked how these costs would change if KASPER changed to require real-time internet-based data submission of CS prescription information, one physician indicated no change while three indicated additional costs would be incurred relative to more personnel time and costs associated with making the technological changes to make software capable of real-time submission. When asked to identify the biggest obstacle to real-time data transmission at the point of CS dispensing, physician dispensers reported that setting up the program software to transmit and the costs associated with those software changes. When asked regarding suggestions on how KASPER could obtain CS data from the practice in real-time, one respondent noted It would work if the computer software people could make a link or submission button for staff to use at check-out. Complete summary tables of survey responses are included in Appendix 7. 15

16 2. Dispensing Veterinarian Responses The majority (61%) of responding veterinarians reported dispensing CS prescriptions one or more times per week whereas nine (35%) respondents reported dispensing CS prescriptions daily. The majority (42%) reported dispensing only between 0-10 CS prescriptions per month, while only 12% of respondents reported dispensing CS prescriptions per month. The vast majority (89%) of dispensing veterinarians that responded to the survey reported they do not request a KASPER report prior to dispensing with only one respondent reporting that a KASPER report is requested prior to each prescription dispensed. Over two-thirds (73%) of dispensing veterinarians indicated they currently submit CS dispensing information to KASPER on a weekly basis while the remaining 7 (27%) indicated varying other time frames of submission, including every 8 10 days, every 9 days, biweekly, and monthly. When asked about internet access at the practice site, all but 5 respondents indicated access is available via desk-top and/or laptop computers. Most dispensing veterinarians submit data on dispensed CS prescriptions by completing the approved paper from and submitting it via fax (65%) or mail (23%) to RelayHealth. Only two (8%) indicated submitting CS data via electronic means. Thus, when asked how workflow would be impacted if KASPER were to begin requiring real-time submission of CS prescription information at the time of dispensing, not surprisingly 16 (62%) indicated it would have a major impact on workflow for a variety of reasons, mainly related to internet access and software issues. A complete list of reasons cited can be found in Appendix 8. When asked about the costs (other than the personnel time required to complete and submit the data) associated with the current method utilized for data reporting to KASPER one-fourth (27%) reported incurring costs such as postage, cost of fax and long-distance charges, paper to copy forms, etc. When asked how these costs would change if KASPER changed to require real-time internet-based data submission of CS prescription information, 10 (40%) of respondents indicated no change in costs would occur whereas 15 (60%) indicated additional costs would be incurred relative to more internet access, increased personnel time and costs associated with making the technological changes to make software capable of real-time submission. When asked to identify the biggest obstacle to real-time data transmission at the point of CS dispensing, veterinarian dispensers reported a variety of obstacles including incompatible software, lack of internet access, disruption of workflow, time, and training. Complete summary tables of survey responses are included in Appendix 8. 16

17 VII. Software Vendor Interviews Seven major vendors providing dispensing system software to pharmacies in Kentucky were identified for potential stakeholder interviews. Vendors were identified based on feedback from the independent pharmacy stakeholder interview and from the chain pharmacy managers interviews. Initial contact with software vendors was made via company websites and through assistance from the ASAP. Once the contact information for appropriate individuals to interview was obtained, s requesting participation in the telephone interviews were sent by the KASPER evaluation team. Positive responses, implying consent to participate in the study, were received from 3 of the 7 vendors identified, including one vendor who offers 9 unique pharmacy management systems. Telephone interviews with the software vendors were conducted during March and April 2011 by members of the KASPER evaluation team. A series of standard questions were presented to the participants. The full set of questions is included in Appendix 9. Combined, the 3 vendors represented 11 pharmacy management systems currently utilized in several hundred pharmacies throughout Kentucky. In addition to the software vendors, a representative for ASAP, which sets the current standard for PMP data transmission also was interviewed. As described previously, the responses to the questions were documented and reviewed for themes that emerged across the 3 interview sessions with the software vendor representatives and the ASAP representative. The major themes that emerged from each of the questions are summarized below. How does your software transmit data to KASPER? Does this transmission require any action on the part of the pharmacists or other person at the individual store level? According to the participants, this varies based on the individual management system in question. Some of the newer systems conduct the reporting in the background, mostly behind the scenes while other older systems require some facilitation by the pharmacists to indicate that the batch file is ready to be submitted. Other pharmacies may be using the automatic extraction process at the point of adjudication through RelayHealth. What changes have you made for other state PMPs? Ex, Oklahoma real-time law? What were the costs associated with these changes? What were the lessons learned? Format changes to keep up with data standards such as those required by NCPDP, ASAP, and individual state PMP requirements, are ongoing. According to one participant the costs associated with keeping up with PMP program requirements has increased. In the past, PMP reporting has been a fairly minor piece of work so it hasn t been carved out separately from other parts of our programs. We are now getting to the point where we are tracking the hours that it takes for functionality. When asked regarding lessons learned, having a national standard common across 17

18 all states PMPs so individual coding does not have to be done for each state and timing were identified as two major lessons learned. Having to individually code for each and every state is very cumbersome and costly from a programming standpoint. Another participant stated The most challenging part has been the timing it has been really bad but it s not their (OK PMP) choice. They have a piece of legislation they have to comply with. The new CMS standards for pharmacy begins January 1, 2012, and this coincides with the deadline for OK s PMP, so the choice for us is between claims (for CMS deadline) and PMP stuff. If real-time transmission of dispensing information is defined as transmission that occurs from the site of dispensing to the KASPER program at the point of sale when the prescription is picked up by the patient for whom it was prescribed. With this definition in mind is your software capable of real time data transmission at the point of controlled substance dispensing? If not, is it capable of real-time transmission at the point of adjudication? How could it be modified to support real time? How much effort or cost would this require? How long would the conversion take? The general consensus was that no software system is currently able to handle CS data transmission to the PMP database real-time at the point of sale. Right now we can do it at the point of adjudication. Another participant stated: None of our systems are currently able to handle it real-time at the point of sale. Issues regarding transmitting data at the point of adjudication or billing were raised. There are a lot of challenges to tying PMP reporting to billing. Those two things are not meant to be done in one step, they are meant to be done in separate steps. There are a lot of caveats of using the switch cash prescriptions don t go through the switch, pharmacy off-line plans (like with a local employer) don t go through, partial fills and long-term-care cycle fills. When asked regarding the effort or cost associated with software modifications to support real time data transmission participants replied It would require significant programming effort, not sure of the costs yet. Another stated I could try to get some more precise cost estimates if needed, but my own gut feel estimate for a given pharmacy system is on the order of the lower tens of thousands. One participant noted We are spending considerably more (to ensure PMP compliance) than in the past and have considered passing these costs along to the pharmacies. When asked how long it would take for conversion of software to support real-time data transmission, participants were unsure. Not sure, it s a fairly big programming effort. Another stated: Once Oklahoma is done we will have a lot of functionality problems solved. OK has the most clear and current standard based on ASAP. Moving to ASAP 4.1 would be even more clear. As long as you aren t on something before ASAP 95 it s not a big deal. What is the greatest obstacle to real time data transmission? All participants suggested time as the greatest obstacle. If the timeline is really short and aggressive it is a problem. It seems like you could require it by the latter half of 2013 and that would be okay. Another stated: Just making the time for 18

19 (PMP) programming vs. the time needed to meet the other state/federal regulations that are due later this year. If Kentucky moves to real time transmission, what do you perceive are the pitfalls for the implementation at the pharmacy? In response to this question, a participant questioned whether KASPER would require point of sale for data transmission or if there would be a real-time process outside of point of sale. Another participant stated: Why real-time? Incomplete data, etc may cause errors. You might make a mistake calling someone a doctor shopper. If you go real time, then you have issues like clerk danger, putting people in bad positions to deny a client a prescription. Any other issues we should consider? Several interesting points were raised in response to this question. First, one participant suggested that the logistics of going real-time would significantly impact workflow. Would CS prescription data go to the PMP and then be accepted by the PMP before being sent to the payer for adjudication? To this individual, it seemed the only advantage to real time is to stop the prescription transaction. If that is not the intent then states can get close to real time by having nightly batches. No special software development would be required and this would considerably reduce the time lag in data for providers who request KASPER reports. States are truncating the time frame for submitting the files. They could send batch files every hour, daily, etc. This is preferable to real time and could be very fast but would not impact the transaction process and workflow. Vendor representatives also expressed appreciation for the outreach efforts being made by CHFS and the KASPER program. Bringing in the vendors like you are, that really helps to make this stuff work. Ask these questions and get a good assessment and give adequate time for the changes that have to occur in the system. Try to keep the playing field as level as possible there has been some concern in Oklahoma that mail order does not have to follow the same rules. VIII. Summary and Recommendations PMPs have been implemented by states as a mechanism to address the growing public health crisis of prescription drug abuse. PMP characteristics vary considerably among the states, including the frequency with which CS prescription data is reported from dispensers to the state PMP databases. While the majority of states require weekly or biweekly reporting, three states require daily reporting (Minnesota, North Dakota and Oklahoma) and one state - Oklahoma - has a mandate effective January 1, 2012 for real-time reporting defined as within 5 minutes of the point-of-sale of a CS prescription at a pharmacy or other dispenser 6. 6 Oklahoma Bureau of Narcotics and Dangerous Drugs Control Prescription Monitoring Program, accessed May 24, 2011, 19

20 Currently, dispensers of CS in Kentucky are required to report CS prescription data to the KASPER program database weekly. To identify the barriers to submission of CS prescription data to the KASPER program in real-time and to gather opinions on the advantages and disadvantages of real-time reporting, multiple approaches were taken, including an interview with the Oklahoma PMP manager, a continuing education program and stakeholder interview with independent pharmacy providers, stakeholder interviews with representatives from retail pharmacy chains, surveys of non-pharmacy dispensers of CS and stakeholder interviews with pharmacy software vendors. The information gleaned from these approaches was used to inform the recommendations outlined herein. At the onset of this study, Oklahoma was identified as the only state currently moving toward real-time data transmission. Dispensers are required to report CS prescription data to the PMP daily at the present time and within 5 minutes of the sale of CS (point of delivery to the patient or patient representative) effective January 1, Overall, it appears from the information gleaned from the Oklahoma PMP program manager that the keys to success for implementing real-time data transmission to the PMP database include involving all stakeholders in the implementation process and providing ample time for planning and preparation before implementing a mandated change to real-time reporting. The continuing education program on KASPER and technology enhancements was well received by the independent pharmacists in the audience who were engaged and interested in providing feedback. Independent pharmacists questioned the definition of real-time (point of sale vs. point of adjudication). Approximately one-third of independent pharmacists in the audience reported not having an integrated point of sale system at the present time. Thus, if KASPER were to define real-time reporting as at the point of sale, the costs associated with purchasing and using integrated point-ofsale systems would be incurred by those pharmacies that currently do not utilize them. Additionally, participants questioned the value of real-time. Specifically, they questioned the role of the pharmacist as gatekeeper vs. the role of the prescriber. Finally, many independent pharmacists who use automatic extraction through RelayHealth as the current method for data reporting are under the misimpression that switching fees are incurred for data submission. This clearly represents an educational opportunity for the Cabinet and KASPER program officials. Similar to the independent pharmacy stakeholders, chain pharmacy representatives questioned the definition of real-time and raised significant issues relative to the impact of real-time reporting at the point of sale on pharmacy workflow and the accuracy of data if real-time reporting occurs at the point of adjudication. Additionally, because chain pharmacies typically have their own proprietary pharmacy management software 20

21 systems, the chain pharmacy managers were also able to provide some perspectives on the costs associated with making the technological changes needed to accomplish real-time data transmission. While the majority did not specifically state a cost, all agreed significant costs would be incurred and one company representative suggested the costs associated with developing and implementing software changes needed for real-time transmission would cost over $100,000. Surveys of non-pharmacy dispensers of CS including physicians and veterinarians gleaned similar information. According to respondents, significant costs would be incurred if real-time data reporting were mandated. Although some currently submit data electronically, the vast majority of non-pharmacy dispensers responding to the survey indicated that data were currently submitted via the required paper form and transmitted via facsimile to RelayHealth. Additionally, several non-pharmacy dispensers reported having no internet access for web-based submission of data should it be required in the future by the KASPER program. Notably, some non-pharmacy dispensers report they currently submit CS prescription data to the KASPER program every 14 days. Considering that current state law mandates that all dispensers of CS report CS prescription information to KASPER every 7 days, it appears that this is an additional educational opportunity for the Cabinet. Finally, interviews with representatives from pharmacy management system software vendors and a representative from the ASAP were conducted and, as with other stakeholders, these participants questioned the value of real-time reporting. Although all agreed the technology is available to implement real-time reporting, the general consensus was that no software system is currently able to handle CS data transmission to the PMP database real-time at the point of sale. Without extensive programming changes, representatives indicated real-time transmission could occur only at the point of adjudication. When considering real-time vs. more frequent batch reporting, one must question the rationale behind the desire for real-time. From a healthcare practitioner who requests KASPER reports for treatment decisions viewpoint, what additional value is gained by seeing real-time data vs. data that is 24 hours old for example? Is the desire of realtime reporting to stop the prescription from being written or is it to stop the prescription from being dispensed? There appears to be differences in the need for and desire of current information when viewed from a regulatory (legal) vs. healthcare provider standpoint. 21

22 Data from an independent evaluation of the KASPER program 7, as well as from the 2010 Satisfaction Survey 8 conducted by KASPER indicate that nearly every user of KASPER believes KASPER is effective at reducing doctor shopping and curbing the abuse and diversion of CS prescriptions. Furthermore, data indicate that when utilized, the information in a KASPER report impacts healthcare treatment decisions. However, in 2009 only 27.5% of prescribers and 16% of dispensers were registered KASPER users. This begs the question from a policy standpoint whether a more costeffective approach to enhancement of KASPER would be finding mechanisms by which the use of KASPER for treatment decisions is increased. The take home message from all stakeholders relative to the biggest obstacle in KASPER moving to real-time was the need to understand why real-time reporting is needed vs. implementing more frequent batch transmission (i.e., daily). Daily batch transmission would serve to improve the currency of information available for prescribers and dispensers to use for treatment decisions at the point of care but would not be associated with any implementation costs for CS dispensers. The Cabinet should consider whether daily transmission, coupled with strategies to increase the use of KASPER by prescribers and dispensers, would be a more cost-effective mandate for all involved than one for real-time reporting. Considering the overall results from the study, one can look at the policy options to reduce abuse and diversion as choices along a program-benefit/program-expense continuum. One can view program expenses in a broad sense consisting of the actual cost to operate the program, the expense for providers and pharmacies, and the expense to patients. Benefits are similar, with program benefits to patients, providers, and the state/public. Ideally, any program should deliver the maximum benefit for the lowest expense: an efficient program. When additional program expenses begin to deliver fewer benefits per expense, policy makers use the term flat-of-the-curve to denote diminishing marginal returns, or the point where additional program expense yields very little increase in benefits. Consider the following hypothetical graph (Figure 4) depicting policy options for the reporting frequency for the KASPER program. The current program based on weekly reporting produces benefits for the current expense. Based on survey data and information from stakeholder interviews, it appears that moving to daily reporting would incur little additional expense and likely yield additional program benefits. 7 Independent Evaluation of the Kentucky All Schedule Prescription Electronic Reporting (KASPER) Program, accessed June 11, 2011, KASPER Satisfaction Survey, accessed June 11,2011, 5EB17933FDDB/0/2010KASPERSatisfactionSurveyExecutiveSummary.pdf 22

23 Figure 4: Benefit-Expense Continuum for PDMP Reporting Frequency Program Benefit Daily Reporting Weekly Reporting Real-Time Reporting No Program Program Expense Moving to real time reporting requires a much greater program expense and likely produces few additional benefits over daily reporting. Again, if the ideal point is locating where benefits of the program are greatest for the lowest expense, our current analysis, taken from the information gleaned from stakeholder interviews and dispenser survey, suggests that moving beyond daily reporting requires considerable new program expense while likely only producing modest gains in benefits. As the Cabinet considers strategies to improve the KASPER program, a range of options beginning with simple program operational changes to increase the effectiveness of the current program to more complex changes focused on increasing use of KASPER reports prior to making prescribing decisions should be considered. The following potential policy modifications are proposed for consideration. Policy Options for Improving the Effectiveness of KASPER to Reduce Prescription Drug Abuse and Diversion: Increase the effectiveness of the current program. Optimize CHFS policy and statutes to allow review of program outliers. A recent review of the KASPER program revealed considerable variation across providers in prescribing CS in Kentucky 9. Most providers issue very few CS prescriptions. For example, the typical KASPER account holder issued an average of 1,665 CS prescriptions in 2009, compared to 250 CS prescriptions issued by non-account holders. Yet the distribution of prescribing is highly skewed; in 2009, 90% of all CS prescribers issued fewer than 2,500 CS prescriptions. 9 Independent Evaluation of the Kentucky All Schedule Prescription Electronic Reporting (KASPER) Program, accessed June 11, 2011, 23

24 If one defines the top and bottom 1% as outliers on either end of the CS prescribing continuum, a review of the top 1% of CS prescribers shows that in 2009 the top prescribers in this bracket issued more than 40,000 CS prescriptions. The Cabinet should work with the medical licensing boards to establish a review process of outlier prescribers to ensure the safety of Kentucky patients. Encourage use of KASPER. From 2002 to 2009 CS prescribing in Kentucky increased by 40% and KASPER system requests increased by 400%. Prescriber use of the KASPER system is growing, but only about 27% of prescribers have registered accounts. The Cabinet should consider options to increase prescriber use of KASPER. o Increase outreach and continuing education effort on the use of KASPER and its usefulness in making informed CS prescribing decisions. o Require all CS prescribers to have KASPER accounts. o Require providers to review KASPER reports for all new CS prescriptions, with mandatory review of KASPER reports every 6 months when prescribing CS for long-term use. o Implement proactive KASPER reporting, where potential outlier use of CS by patients triggers a KASPER alert sent to prescribers. Improve the operation of the current program. With more prescribers using the KASPER program, the efficiency of the program could be improved by moving from weekly reporting to daily reporting at little to no cost to dispensers. 24

25 Appendix 1: Slide Set for Continuing Education Program and Independent Pharmacy Stakeholder Interview 25

26 26

27 27

28 28

29 29

30 30

31 31

32 32

33 Appendix 2: RelayHealth Instructions for Submitting Electronic Batch Files 33

34 34

35 Appendix 3: RelayHealth Instructions for Automatic Extraction of Prescription Data 35

36 36