Maine’s Prescription Drug Monitoring Program: Evaluation Report

I. Introduction

Maine’s Prescription Drug Monitoring Program (PMP) was implemented in July 2004 with federal funding to help curb an alarming rise in the illicit use of prescription drugs.[i] Under the program, all transactions from pharmacies dispensing prescriptions of Schedules II, III, and IV drugs are submitted electronically to a database, maintained by the Maine Office of Substance Abuse. This database is analyzed and used to issue threshold reports (sent to the clinicians who show up as ”prescribers” of the prescriptions) indicating a potential “red flag” on individuals who may be receiving dangerous levels of prescription drugs. Clinicians may also query the database to request a patient history report on a patient to whom they have, or are considering, writing a prescription. This information allows clinicians and pharmacies to better administer prescription drugs to limit and curb the dangerous and deadly effects of abuse and overdose.

In developing the Maine PMP, state policymakers and stakeholders identified a strong preference that the program be used as a public health and clinical intervention tool to reduce the illicit use of prescription drugs and not be used as a law enforcement tool, as it is in a number of other states with a PMP funded under the Harold Rogers Program. The strong public health orientation of Maine’s PMP is reflected in the program’s goals, which are to:

· curb illicit use of prescription drugs in Maine;

· give prescribers an added tool in patient care;

· get patients who are addicted into proper treatment;

· help reduce prescription drug overdoses;

· ensure that those who need strong prescription drugs receive them.

The Muskie School, University of Southern Maine, conducted a study of the implementation of Maine’s PMP Program in 2005 (Lambert 2006). The evaluation found that the PMP had been successfully implemented and was working well so far:

· Prescribers receiving Threshold Reports and requesting Patient History Reports had used them to help clarify whether patients were “doctor shopping” or using prescription medications appropriately.

· Prescribers and dispensers were enthusiastic about the planned availability of an on-line web-portal (scheduled for implementation in early 2006) that would allow close to “real-time” access to information.

· Early concerns over patient confidentiality, the potential use of PMP data by law enforcement, and a potential “chilling effect“ (in which concerns over confidentiality would constrain the number of prescriptions written) had not materialized.

· Major stakeholders (including The Maine Medical Association and the Maine Osteopathic Association) were pleased with how the PMP had developed and optimistic about what it might accomplish.

The study noted that the PMP would need to continue to increase the number of registered prescribers actively using the program and its database; maintain its exemplary record of data security and confidentiality; reduce the time between when information was requested and received; and maintain and enhance the public health function of the PMP. The study, incorporating the advice of PMP policymakers and stakeholders, recommended that the PMP program begin to consider longer-term issues of sustainability and how the impact of the program might be monitored and assessed over time.

The Maine Office of Substance Abuse contracted with the Muskie School to conduct a study of the progress and impact of the PMP following its implementation. This study, presented in this report, was designed to provide feedback to PMP policymakers and stakeholders about what the PMP had accomplished in its first two years and what it might be expected to accomplish in the next few years in terms of reducing the diversion or illegal use of prescription drugs and the consequence of this use. The opportunity – and challenge of this study – was to be concrete in measuring the progress and impact of the PMP, but to be realistic in terms of what these impacts might be, given that the program is still relatively new and growing.

II. Context and Scope of Study

An alarming increase in the abuse of prescription drugs in Maine prompted state policymakers to develop Maine's Prescription Monitoring Program (PMP). Treatment admissions for prescription drug abuse had increased from 83 in 1995 to 1148 in 2003. The number of overdose deaths increased steadily – as did the proportion of these deaths caused by prescription drug abuse. In 2001 there were 90 drug deaths in the state; 70 (78 percent) were caused by a pharmaceutical. One year later, in 2002, the number of overdose deaths had nearly doubled to166; 148 of these deaths (89 percent) were caused by a pharmaceutical. Arrests for prescription drug diversion increased steadily, accounting for 16 percent of arrests made by Maine Drug Enforcement Agency in 2003. In 2002 more than twenty percent of Maine high school seniors reported that they had used prescription drugs to get high.

Prescription Drug Monitoring Programs in other states, funded by the U.S. Department of Justice, offered an opportunity and a model to address Maine’s growing prescription drug abuse problem. The first Bill to create a Prescription Monitoring Program was introduced in the Maine Legislature in 2001. While there was growing recognition that Maine had a prescription drug problem that needed to be addressed, there was concern that the program should not be used as a tool for law enforcement – as it is in a number of other states. A related concern was that the data collected and used needed to be confidential and secure and be in compliance with emerging HIPAA regulations.

A consensus emerged that a prescription monitoring program should be used as a public health and clinical intervention tool to reduce the illicit use of prescription drugs.

Under the leadership of Maine’s Office of Substance Abuse and with the participation and support of Maine’s medical community, pharmacies, attorney general’s office, department of licensure and regulation, and other stakeholders a working consensus was formed for how Maine’s Prescription Monitoring Program should work to support this goal. The passage of the Bill in 2003 (on the third try before the Maine Legislature) creating the Prescription Monitoring Program, gave the Office of Substance Abuse the authority to develop the program, but did not authorize a state expenditure. To be implemented, the program would need to secure external funding, which it did in October 2003. Many data confidentiality issues were addressed in the enabling legislation creating the PMP.[ii] In administering the PMP, OSA is designated as a “health oversight agency” under HIPAA.

The implementation of the PMP proceeded smoothly, with 350 prescribers and 66 dispensers registering for the program by summer 2005 (Lambert 2006). Prescribers received and used threshold reports and requested and used patient history reports to monitor patients’ use of prescription drugs. Data confidentiality had been maintained in an exemplary manner. The main concern of (and recommendation by) prescribers and dispensers was to have access to more “real time” information from the PMP database, which would allow for proactive management for patients. The launch of OSA’s WEB Portal, planned for the first quarter of 2006, held much promise to provide such improved access. For the PMP to be able to meet its longer term goals of reducing the abuse of prescription drugs, and the consequence of this abuse, more prescribers would need to continue to register for and use the PMP database. What is not known is how many and what type(s) of prescribers among the 6,139 Drug Enforcement Agency registrants in Maine would need to participate actively in the PMP to reduce prescription abuse rates and overdoses. Many professional and mid-level licensed prescribers are not likely to prescribe Schedule II, III, or IV prescriptions. One may conjecture that the “tipping point” would be a sizeable portion, if not a majority, of the DEA registrants, particularly primary care and emergency department clinicians, who may be particularly likely to see new patients requesting prescriptions to control pain.

This study examines the following questions to see if the PMP has begun to achieve the impacts, which if sustained, are likely to result in longer term outcomes of reducing prescription drug abuse and overdoses:

Following its implementation, has the PMP been expanded and refined as planned, particularly with respect to recruitment and participation of dispensers and prescribers and how the PMP database is used?

Which prescribers are using the PMP? Is the PMP growing in the regions of the state where it is most needed?

Has the PMP given prescribers a useful tool in patient care? When and how can this trend be related to the composition and growth of the PMP?

Has patient care improved as a result of the PMP, with respect to: pain management, identification and treatment of substance abuse problems, prescribing of medications in general?

What are the collateral effects of the PMP on other programs and regulatory activities in Maine?

Have there been any adverse or other unintended consequences of the PMP [e.g. chilling effect]?

Has the abuse of prescription drugs in Maine changed over time? Has the number of drug overdoses involving prescription drugs changed? Has the number of admissions for addiction changed? When and how can these trends be related to the composition and growth of the PMP?

What data collection and evaluation activities will be needed in the future to assess the long-term impact of the PMP?

III. Methods and Approach

This impact evaluation is based on data from four sources:

Survey of prescribers who have registered in the PMP system.
Survey of dispensers who submit data to the program.
Key stakeholder interviews with OSA staff, members of the PMP Advisory and Clinical Advisory Committees, and heads of professional licensing boards.
Secondary data analysis of standard and special reports, queries from the PMP database and compilation of aggregate data trends from the public-use databases.

Prescriber and Dispenser Surveys: Surveys were mailed in August 2006 to all prescribers and to all dispensers who had registered for the PMP. A second mailing of the surveys was sent out in October 2006 to prescribers and dispensers not responding to the initial mailing, or who had registered for the program since the first mailing. A total of 354 out of 968 prescribers (36.6 percent) and 34 out of 102 dispensers (33.3 percent) mailed back completed surveys. The response rates for the prescriber and the dispenser surveys are similar to those obtained in the surveys conducted for the implementation evaluation (38.9 percent and 31.8 percent respectively). The geographic locations (county) and medical specialties of prescribers responding to our survey are very similar to the geographic and specialty distributions of all prescribers registered for the PMP database. We are confident that the answers from prescribers are reflective of registered prescribers throughout Maine.

Stakeholder Surveys: Interviews were conducted with PMP staff, advisory and medical committee staff, contractors, and members of health professional licensing boards to gain additional perspective on how the program was developing, and whether, when, and how one might expect the PMP to impact the rate and consequence of prescription drug abuse. Interviews with heads of licensing boards and other agencies explored whether they were receiving a collateral benefit from the PMP Program and what the potential for such benefits in the future might be. Information from the stakeholder interviews is incorporated or noted, where appropriate, in different sections of the report. The names of stakeholders are withheld to protect confidentiality.

Secondary Data Analysis: The PMP database was queried to generate information on the number and distribution of threshold and patient history reports. To protect confidentiality, these queries were conducted by PMP staff, at the request of the researchers of this report. Public health databases in the public domain were also queried to report trend information on prescription drug abuse and its consequences in Maine.

Approach: As the Maine PMP grows and matures, it should reduce the prevalence and consequence of prescription drug abuse in Maine, or at least curb a growing trend. Ideally, an outcome evaluation of the PMP should be conducted several years after it has achieved a critical threshold of participants. This study was conducted following a very successful implementation and start-up, but before the PMP reached a size and maturity in which it might be expected to achieve its longer-term goals and outcomes. Consequently, the approach we take is to assess whether the major components and strategies of the PMP program - including outreach, recruitment, and active participation of prescribers and dispensers - are developing as planned and in specialty areas and clinical venues to be able to achieve the desired outcomes. If these processes are occurring, the PMP will achieve intermediate level outcomes, including increased discussion and consultation among prescribers, dispensers, and patients; increased substance abuse education, prevention and treatment; and collateral benefit to other agencies in Maine. If the PMP can achieve these intermediate outcomes, then longer term outcomes may result in the future, including reduced overdoses, admissions for addiction, and use of illicit drugs.

IV. Findings

Expansion and Growth of the PMP Program

Central to the growth and success of the PMP Program is the active involvement of prescribers and dispensers in using the data available from the program. Prescribers receive data about their patients in one of two ways: (1) through a Threshold Report sent to them by the PMP Program indicating that a patient has a “suspicious number of prescriptions filled in a certain time period”; (2) by requesting a Patient History Report on one of their patients. To be able to request a patient history report, a prescriber must register with the PMP Program. Dispensers are not sent a Threshold Report (although notified prescribers may contact them about a particular patient), but may register and request Patient History Reports.

At the time the implementation study was conducted in summer 2005 –one year after the PMP program was implemented – 350 prescribers had registered to use the PMP database. The vast majority of these prescribers were familiar or very familiar with the program; the most common way they had learned about the program was through a mailing (40 percent), an information pamphlet (24 percent), or a professional association (21 percent). Just under half had requested a Patient History Report; sixty-one percent not having requested a patient history report expected to do so within the next six months. Most prescribers receiving threshold and patient history reports found them useful and had been able to clarify whether or not their patients were using prescriptions properly or improperly. The major issue prescribers had with the system was being able to access data on a more timely or “real-time” basis when the patient was in the clinical setting. To meet its longer term goals, the PMP will need to continue to increase the number of registered prescribers actively using the program. It is important, as the program matures, for PMP policymakers and stakeholders to have a better sense of which types (specialties) of prescribers are registering and using the program.

Outreach to Prescribers and Dispensers: Since January 1, 2004, the Maine Office of Substance Abuse has provided 40 trainings to 1,398 participants (Table 1). Trainings to dispensers occurred in the start-up of the program, while trainings to licensing and law enforcement stakeholders occurred a bit later. The largest number of trainings (27) has been provided, on an ongoing basis, to prescribers. This reflects the need to continue to promote the PMP program and to increase the number of active participants. Stakeholder interviews conducted with medical associations and members of the PMP advisory and clinical committees strongly suggest that outreach trainings were well received and considered helpful and informative.

Table 1. Number of PMP trainings provided by OSA, January 1, 2004-September 30, 2006 (number of participants in parenthesis).

Period	Dispenser Trainings	Prescriber Trainings	Licensing	Law Enforce- ment	Combined	Total
1/1/04 – 6/30/04	2 (135)	3 (180)	0 (0)	0 (0)	0 (0)	5 (315)
7/1/04 – 12/31/04	2 (150)	4 (100)	0 (0)	1 (35)	0 (0)	7 (285)
1/1/05 – 6/30/05	1 (100)	4 (145)	3 (24)	1 (40)	0 (0)	9 (309)
7/1/05 – 12/31/05	0 (0)	6 (143)	1 (8)	0 (0)	1 (8)	8 (159)
1/1/06 – 6/30/06	0 (0)	8 (282)	0 (0)	1 (30)	0 (0)	9 (312)
7/1/06 – 9/30/06	0 (0)	2 (18)	0 (0)	0 (0)	0 (0)	2 (18)
TOTAL	5 (385)	27 (868)	4 (32)	3 (105)	1 (8)	40 (1,398)

Prescribers Registering for and Using the PMP Program

The number of providers registering for the PMP has increased steadily over time (Table 2), paralleling OSA’s educational and outreach activities to this group. As of September 30, 2006, nearly 1,000 prescribers in Maine had registered for the PMP. The largest increase in new registrants - 49.2 percent – occurred between January 1 and June 30, 2006. This is the period when OSA’s much anticipated WEB Portal came on line, in March 2006.

Table 2. Number of Prescribers Registered for PMP Program,

January 1, 2005-December 30, 2006

Period	Number of prescribers Percent registering Increase
January 1, 2005 – June 30, 2005	309
July 1, 2005 – December 31, 2005	181 36.9%
January 1 2006 – June 30, 2006	317 49.2%
July 1, 2006 – September 30, 2006*	294 48.1%
TOTAL	1,101

Important questions for understanding how the PMP is working now, and what impact it might have in the future, include what type of prescribers are registering for the program and from which areas of the state. We used public licensure lists to assign primary specialty designations to prescribers who had registered for the PMP Program by August 2006 (Table 6). It was not possible to assign specific specialties to 165 of the 900 (18.3 percent) of the prescribers registered by August 2006 (Table 3). Even with this limitation, the information in Table 3 is revealing, showing that the five largest categories of specialists registering for the PMP are family physicians (n=304), mid-level practitioners (n=134), internal medicine (n=90), psychiatry (n=55), and emergency medicine (n=49). These are the specialties one would expect to have the most interest and need for using the PMP database. The relatively high number of psychiatrists who have registered is a bit of a surprise.

In general, prescribers are registering for the PMP throughout the state, proportionate to the population of their respective counties (Table 4). Possible exceptions include Oxford and York counties, which have fewer registered prescribers, relative to their population, than other counties in Maine. It may be that a significant number of residents in York County may have received prescriptions from out-of-state doctors (New Hampshire and Massachusetts), as well from prescribers in Cumberland County. That this may be happening is suggested by York County having the lowest script / population ratio in the state (1.31 vs. state average of 1.50). Because a person may receive their prescription(s) from clinicians outside of their county, region may be a better level at which to examine this relationship. As shown at the bottom of Table 4, the number of registered prescribers is consistent with the population size within each of Maine’s three major regions.

Table 3. Registered Prescribers in Maine, August 2006,

by specialty

Specialty	n
Anesthesiology	5
Mid-level	134
Emergency Medicine	49
Gastroenterology	1
Family Medicine	304
Internal Med.	90
Neurology	7
Neuromusculoskeletal	2
OB/GYN	13
Occupational Med.	8
Oncology	1
Orthopedics	3
Osteopathic Manipulative Medicine	3
Phys Medicine/Rehab	16
Podiatry	2
Psychiatry	55
Pulmonary Medicine	3
Surgery, Orthopedic	22
Surgery, Other	16
Not Available	166
TOTAL	900

It would be useful to be able to assess whether participation (number of registered prescribers and users) in the PMP program is growing in the areas of the state where it is most needed (measured in terms of diverted prescription drugs or adverse events including overdoses and deaths). The data and method by which to examine this question are not currently readily available. (How this might be done is taken up in the recommendation section of this Report). We have included the ratio of scripts per person in the last column of Table 4 as a crude proxy for need to monitor potential abuse of prescription drugs. This is a crude proxy because these ratios may reflect demographic and related illness factors. With this caveat in mind, we note that the counties with the highest ratios of scripts per capita are Aroostook (1.58) Knox (1.64) Washington (1.66), and Penobscot (1.74). Region III (Aroostook, Hancock, Penobscot, Piscataquis, Washington Counties) has the highest ratio (1.65). The distribution of registered prescribers by specialty, by county, is presented in Appendix Table 1.

Table 4. Number of registered prescribers by county and region; and county and scripts per person.

County	Registered Prescribers* n (percent)	Population n (percent)	Scripts(Schedule II, III, IV) per person
Androscoggin	109 (11.0)	107,022 (8.1)	1.44
Aroostook	66 (6.7)	73,390 (5.6)	1.58
Cumberland	231 (23.3)	273,505 (20.8)	1.45
Franklin	17 (1.7)	29,736 (2.3)	1.32
Hancock	42 (4.2)	53,556 (4.1)	1.57
Kennebec	129 (13.0)	120,645 (9.2)	1.55
Knox	24 (2.4)	41,008 (3.1)	1.64
Lincoln	54 (5.5)	35,236 (2.7)	1.54
Oxford	16 (1.6)	56,614 (4.3)	1.49
Penobscot	139 (14.0)	148,196 (11.3)	1.74
Piscataquis	18 (1.8)	17,525 (1.3)	1.56
Sagadahoc	9 (0.9)	36,927 (2.8)	1.39
Somerset	37 (3.7)	51,584 (3.9)	1.58
Waldo	17 (1.7)	38,392 (2.9)	1.56
Washington	19 (1.9)	33,558 (2.5)	1.66
York	63 (6.3)	200,359 (15.2)	1.31
Region I (a)	294 (29.7)	473,864 (36.0)	1.39
Region II (b)	412 (41.6)	517,164 (39.3)	1.51
Region III (c)	284 (28.7)	326,225 (24.7)	1.65
TOTAL	990 (100.0)	1,317,253 (100.0)	1.50 (state average)

a. Cumberland, York

b. Androscoggin, Franklin, Kennebec, Knox, Lincoln, Oxford, Sagadahoc, Somerset, Waldo

c. Aroostook, Hancock, Penobscot, Piscataquis, Washington

Patient Care

To find out more about whether and how registered prescribers are actively using the PMP Database, we surveyed all prescribers having registered for the database by September 2006. Thirty-seven percent of all registered prescribers completed the survey. This survey, conducted one year after the first prescriber survey was administered, updates our knowledge of which prescribers are using the PMP and how they are using it. Seven out of ten providers reported that they were either familiar (41.7 percent) or very familiar (28.7 percent) with the PMP program. This is a bit less than the nine out of ten prescribers reporting that were either familiar or very familiar with the program in the implementation survey one year earlier. The number of registered prescribers has nearly tripled in the past year; it may be that earlier registrants tended to be more familiar with the PMP before they registered for it.

Many prescribers had learned about the PMP program through mailing of information (25 percent) and through their professional association (24 percent) (Table 5), which were reported to be important sources of information by the prescribers surveyed a year earlier. However, the most common way that the prescribers reported they learned about their program was through their colleagues or employers (30 percent). This is consistent with the idea held by PMP stakeholders that a strong “selling point” for the program is how useful prescribers find it and their willingness to recommend it to colleagues. This suggests that the PMP program is likely to continue to grow, as more prescribers come to use it and to recommend it to their colleagues.

Table 5. How Prescribers learned about Maine’s Prescription Monitoring Program

(sources of information are not mutually exclusive; n=354)

How did you learn about your responsibilities and rights under the PMP? (check as many as apply)	n
Professional Association	84
Information Pamphlet	44
PMP Website	40
Mailing	89
Training Session	33
Threshold Reports	51
Colleague/ employer	104
Other	4
Not familiar with program / NA	17

Threshold Reports: Three out of four prescribers (75 percent) reported that they had received a threshold report and the vast majority found the report easy to understand (98 percent) and helpful (94 percent). These results are similar to, and even more positive than, those from our earlier survey of prescribers, where 71 percent reported having received a threshold report; 94 percent reported them to be easy to understand and 80 percent found them to be useful. The increase in those describing the reports as useful may reflect greater familiarity and experience with the reports.

Indeed, prescibers in our second survey receiving a threshold report are much more likely to report having used that information in the active management of their patient than prescribers in our first survey (Table 6). The majority (64 percent) of prescribers surveyed in 2006 receiving a threshold report, entered information from that report in a patient’s record and/or spoke with the patient. Over half of these prescribers (57 percent) were able to confirm that at least some patients were not misusing prescriptions; 42 percent were able to confirm that at least some of their patients were abusing prescriptions. The results reported in Table 6 strongly suggest that prescribers are using threshold reports in the way designers of the PMP intended.

Table 6. What happened as a result of prescriber receiving threshold report on one or more patients (responses not mutually exclusive), 2005(n=97) and 2006 (n=262) Prescriber Surveys

	2005 Survey N (%)	2006 Survey N (%)
Placed Information in patient’s record; spoke with patient	10 (10.3)	169 (63.5)
Contacted other providers to coordinate care	10 (10.3)	94 (35.3)
Confirmed patient not misusing prescriptions	9 (9.3)	152 (57.1)
Confirmed patient was misusing prescriptions	8 (8.2)	112 (42.1)
Reduced / eliminated prescriptions for patients	12 (12.4)	90 (33.8)
Entered into a contract with patient	*	64 (24.1)
Referred / recommended substance abuse treatment for patient	4 (4.1)	54 (20.3)
Referred / recommended pain management for patient	*	44 (16.5)
Dismissed patient from practice	5 (5.2)	28 (10.5)
Nothing	12 (12.4)	19 (7.1)
Other	3 (3.1)	9 (3.4)
DK /NA	32 (33.0)	15 (5.6)

* response category not included in question in 2005 Survey

Patient History Reports: Patient History Reports provide a tool for prescribers to manage proactively their patient’s use of prescription drugs and their healthcare. The implementation study found that just under half of the registered prescribers (45.6 percent) had requested a patient history report and sixty-one percent of those not having requested a patient history report, intended to within the next six months. Prescribers requesting a patient history report used the reports much as they used the threshold reports. While prescribers generally found the patient history reports helpful (76 percent), they strongly recommended that the reports be timelier, so that the information was available close to, or in, “real time” when the patient was in the clinical setting.

The current prescriber survey was mailed in August 2006 (and a second mailing in October) six months after the much anticipated OSA Web portal had come on-line, enhancing the accessibility and timeliness of information to prescribers. The number of registered prescribers increased significantly in the six-month period (January 1 – June 30, 2006) when the WEB Portal came on-line, (Table 2) suggesting that many prescribers returning the survey were new users of the PMP who did not have experience with the PMP before the WEB Portal was implemented.

Thirty-nine percent of the prescribers surveyed reported that they had requested a Patient History Report before the OSA WEB portal was available (Table 7). Most prescribers (62 percent) had requested a report for ten or fewer patients; twelve percent had requested a patient history report for between eleven and twenty patients. Just under half (48 percent) of the prescribers reported that they tended to request patient history reports for established patients; twenty-five percent reported that they tended to request reports for new patients; and nine percent reported that they requested reports for both established and new patients.

Prescribers requesting Patient History Reports before the OSA WEB Portal was available, used this information actively to help manage their patients’ care (Table 8). Three out of four prescribers entered information from the Reports in patients’ records or spoke with the patient about the information; the information was used in equal measure to confirm that patients were (54 percent) and were not (55 percent) misusing prescriptions. Overall, prescribers are using the Patient History Report much in the same way as they are using Threshold Reports.

Patient History Reports From the OSA WEB Portal: The majority of prescribers (55 percent) surveyed report that they have requested and accessed patient information from OSA’s WEB Portal (Table 9). The number of patients for whom prescribers have requested patient history reports is similar to (but slightly higher than) prescribers requesting reports before the availability of the WEB Portal. We would expect this number to increase as prescribers become more familiar with the WEB Portal. Prescribers using the OSA WEB Portal are more likely to request patient history reports on new patients (41.3 percent) than prescribers requesting patient history reports before the availability of the WEB Portal (25.9 percent). This is precisely the use of the WEB Portal envisioned by PMP policymakers and stakeholders – the ability to access timely information on patients not well known to the prescriber. The most common reason (29 percent) given for not having used the WEB Portal, was that the prescriber did not know about it. Since many of the prescribers are newly registered, it may take a little more time and outreach for them to learn about and to use the WEB Portal.

Table 7. Prescribers’ requesting a Patient History Report before OSA WEB Portal was available in March 2006

Before the OSA WEB portal was available, did you request a Patient History Report about the prescriptions filled by one or more of your patients?

Yes

DK/NA

Total

Percent (n)

39.0 (138)

58.8 (208)

2.2 (8)

100.0 (354)

How many patients did you request information about?

1-10

11-20

21-50

51-100

100<

DK/NA

Total

61.6 (85)

11.6 (16)

4.4 (6)

0.0 (0)

0.7 (1)

21.7 (30)

100.0 (138)

Did these patients tend to be:

Established

New

Both established and New

DK / NA

Total

47.8 (66)

25.4 (35)

18.1 (25)

8.7 (12)

100.0 (138)

Table 8. How prescribers used patient history report (one or more patients) before OSA WEB Portal was available (responses not mutually exclusive, n= 135)

	N (percent)
Placed Information in patient’s record; spoke with patient	102 (73.9)
Contacted other providers, pharmacies	62 (44.9)
Confirmed patient not misusing prescriptions	76 (55.1)
Confirmed patient was doctor shopping	75 (54.3)
Reduced / eliminated prescriptions for patients	60 (43.5)
Dismissed patient from practice	37 (26.8)
Referred / recommended substance abuse treatment for patient	36 (26.1)
Referred / recommended pain management for patient	29 (21.0)
Nothing	3 (2.2)
Other	9 (6.5)
DK / NA	9 (6.5)

Prescribers are using information from the OSA WEB Portal in a similar manner to how they used information from Patient History Reports before the WEB Portal was operational (Table10). It is encouraging that over half the prescribers (64 percent) using

Table 9. Prescribers’ using OSA WEB Portal to request and access patient information

Have you used the WEB Portal to request patient information?

Yes

DK/NA

Total

Percent (n)

55.4 (196)

42.4 (150)

2.2 (8)

100.0 (354)

How many patients did you request information about in the past three months? (n=190)

1-10

11-20

21-50

51-100

100<

DK/NA

Total

62.8 (123)

14.3 (28)

7.1 (14)

3.1 (6)

2.0 (4)

10.7 (21)

100.0 (196)

Did these patients tend to be: (n=190)

Established

New

Both established and New

Total

30.1 (59)

41.3 (81)

23.5 (46)

5.1 (10)

100.0 (196)

FOR THOSE NOT USING OSA WEB PORTAL

What is the primary reason you have NOT requested a patient history report)? (n=150; answers not mutually exclusive)

Process too time consuming

Did not know about this aspect of the PMP

Not viewed as necessary

Information not available in a timely basis

Other

No Answer

16.0 (24)

28.7 (43)

19.3 (29)

2.6 (5)

16.0 (24)

29.3 (44)

the WEB Portal were able to confirm that at least some of their patients were not misusing prescriptions, and nearly half (47 percent) were able to confirm that at least some of their patients were misusing prescriptions. Over a quarter of prescribers using the Web Portal referred at least some of their patients for substance abuse treatment (28 percent) or for pain management (26 percent).

Table 10. How prescribers used patient information from OSA WEB Portal (responses not mutually exclusive; n = 196)

	N (percent)
Placed Information in patient’s record; spoke with patient	136 (69.4)
Contacted other providers, pharmacies	77 (39.3)
Confirmed patient not misusing prescriptions	124 (63.3)
Confirmed patient was doctor shopping	93 (47.4)
Reduced / eliminated prescriptions for patients	88 (44.9)
Dismissed patient from practice	31 (15.8)
Referred / recommended substance abuse treatment for patient	54 (27.6)
Referred / recommended pain management for patient	51 (26.0)
Nothing	6 (3.1)
Other	10 (5.1)
DK /NA	9 (4.6)

Prescribing Controlled Substances and Managing Patient Care: Prescribers were asked several questions designed to elicit whether and how the PMP is helping clinicians to better manage patients with pain. First, they were asked a set of questions about how useful the PMP was in helping clinicians and pharmacies, in general (Table11). Next, prescribers were asked whether the PMP has changed the amount of controlled substances they prescribe. Finally, prescribers were asked two open-ended questions about how the PMP has changed the way they help patients manage their pain and how they prescribe medications in general. The open-ended questions were designed to explore whether the PMP program was having a broader impact (beyond controlled substances) on how prescribers managed their patient’s medication.

Table 11. How useful Prescribers find the PMP (n=354)

How useful is the PMP in helping …

Very Useful

Percent

(n)

Useful

Percent (n)

Somewhat Useful

Percent

(n)

Not

Useful

Percent (n)

No answer

Percent

(n)

clinicians and pharmacies to monitor patients’ controlled substance prescriptions?

54.5

(193)

29.1

(103)

7.1

(25)

0.8

(3)

8.5

(30)

to control “doctor shopping” by patients seeking to access /abuse controlled substances?

53.1

(188)

26.0

(92)

6.5

(23)

1.7

(6)

12.7

(45)

clinicians consult with each other about possible prescription abuse by patients?

41.2

(146)

31.9

(113)

12.2

(43)

1.1

(4)

13.6

(48)

The majority of prescribers found the PMP to be very useful in helping clinicians and pharmacies to monitor patients prescriptions (55 percent) and to control “doctor shopping” (53 percent); somewhat less than half (forty-one percent) found the PMP very useful in helping clinicians consult with each other. Forty-one percent of the prescribers surveyed reported that the availability of the PMP changed the way or the amount of controlled substances they prescribed (data not shown). Twenty-one percent reported that they prescribed about the same; eight percent reported that they prescribed more, and sixty-five percent reported that they prescribed fewer controlled substances than two years ago. The majority of the respondents either did not answer, or did not offer detailed answers, to the open-ended questions of whether, and how, they had changed their prescribing practices (for controlled substances and for all medications). In reviewing the open-ended responses that were given, it appears that most respondents felt that changes in prescribing controlled substances had already been captured in earlier questions and that the PMP had “not yet” changed broader prescribing patterns and patient management beyond controlled substances.

How Dispensers Use the PMP: A survey was mailed to 102 registered dispensers; 34 returned completed surveys (33.3 percent return rate). The vast majority reported that they could access the PMP program over the internet (85 percent) and that they had requested a patient history report on one or more of their clients. Just over half the dispensers (52 percent) returning the survey reported that they had requested a patient history report for between 1-4 clients; another 24 percent had requested a patient history report for between 5-10 clients (Table 12). Just over half of dispensers requesting a patient history report said that they had used this information in some cases to confirm that the customer was misusing prescriptions (52 percent) and used it in other cases to confirm that the customer was not misusing prescriptions (55 percent) (Table 13). Thirty percent of dispensers report that in some cases they refused to fill a prescription on the basis of information received.

Table 12. Number of clients for whom Dispensers requested Patient History Reports, in past three months (n=29)

How many customers have you requested information about in the past 3 months?	n (percent)
0	4 (13.8)
1-4	15 (51.7)
5-10	7 (24.1)
11 or more	3 (10.2) (high of 40)

Table 13. How Dispensers have used information from

Patient History Reports (responses not mutually exclusive, n=29)

	n (percent )
Contacted, spoke with prescribers or other Pharmacies	16 (55.2)
Confirmed that customer was misusing prescriptions	15 (51.7)
Confirmed that customer was not misusing prescriptions	16 (55.2)
Refused to fill prescription	9 (31.0)
Nothing	2 (6.9)

Dispensers were asked the same three questions, as prescribers were asked; regarding how useful the PMP program is (Table 14). Dispensers completing the survey tended to find the PMP even more useful than the prescribers.

Table 14. How useful Dispensers find the PMP (n=34)

How useful is the PMP in helping …

Very Useful

Percent

(n)

Useful

Percent (n)

Somewhat Useful

Percent

(n)

Not

Useful

Percent (n)

No answer

Percent

(n)

clinicians and pharmacies to monitor patients’ controlled substance prescriptions?

47.1

(16)

35.3

(12)

11.8

(4)

2.9

(1)

2.9

(1)

to control “doctor shopping” by patients seeking to access /abuse controlled substances?

55.9

(19)

20.6

(7)

17.6

(6)

2.9

(1)

2.9

(1)

Clinicians and pharmacies consult with each other about possible prescription abuse by patients?

55.9

(19)

20.6

(7)

17.6

(6)

2.9

(1)

2.9

(1)

Pharmacies in Maine are required to submit their relevant prescription information to the PMP data contractor. Most dispensers had earlier contact with the PMP than most prescribers in the state. As the glitches in reporting requirements and formats have been worked, many dispensers have been less actively concerned, than in the planning for and early start-up, of the PMP. The information from the dispenser survey reported on in this section, suggests that dispensers are satisfied with the PMP and find it a useful tool. However, we would have a more complete picture of what dispensers think about and how they use the PMP if we had information from the two-thirds of dispensers not returning the survey.

Collateral Effects, Broader Impacts, and Unintended Consequences

The Prescription Drug Monitoring Program has significant potential to benefit other agencies and regulatory bodies in Maine. The program also has the potential to have unintended, but harmful effects, such as decreasing the prescribing of medication needed to control pain (‘chilling effect”), or if a breach of confidentiality over data should occur. The PMP has sustained an exemplary record of maintaining the confidentiality of the data and a chilling effect has not occurred. This success – crucial to the integrity, continuation, and growth of the program – may have also limited some of the potential collateral benefit of the PMP to other programs in the state, such as Drug Courts or the Maine CDC. PMP staff and stakeholders interviewed reported that, by and large, the program has stayed focused on its primary goals and the need to maintain strict data confidentiality, which given lean staffing and resources, has slowed development of working with other state health and social service agencies.

The PMP has maintained very productive, but confidential, relations with state medical licensing boards. If a licensing board wants information about a member, there must be a formal, notarized request. Licensing board directors and PMP staff report that there have generally been only a relatively few number of requests for information from each Board and that these requests have been meet professionally, discretely, and in a timely manner. Law enforcement is only permitted to request information if they have a court order for information that pertains to a specific case. The number of such requests has been very limited and these requests have proceeded smoothly.

As the PMP program grows it provides clinicians a useful tool to help patients to manage their pain and educate clinicians more broadly in pain management and the prescribing of different types of medication. The more clinicians that participate in the program, the more are likely to learn about it, view it in favorable terms, and to join. The data presented above and stakeholder interviews suggest that the PMP is approaching a size in which these broader impacts may be realized within the next few years.

Changes in the Abuse and Consequences of Prescription Drugs

A dramatic increase in Maine in the prevalence of prescription drug abuse, arrests for prescription drug diversion, and drug-related deaths starting in the mid to late 1990s helped create the public support for the PMP Program. The program was created to provide a public health tool to combat the problem and illness of prescription drug abuse. Ideally, one would want to be able to relate the growth and activities of Maine’s PMP to selected indicators related to prescription drug abuse, particularly arrests, treatment admissions, and overdose deaths. However, it is very difficult to do this. First, Maine’s PMP is still relatively new, is growing, and has yet to reach a “mature” size. There will also be a time-lag between when a mature program might reasonably be expected to affect outcome indicators in the aggregate. There are inevitably other events and factors, other than the PMP program itself, that might influence these indicators (in either direction), making it difficult to isolate the effect off the PMP itself, with confidence.

Rather than throw our hands up in the air and walk away from trend data on prescription abuse and its consequences, it is very important to keep these trends– state and national - in mind as the PMP program continues to grow. A recent report by SAMHSA, “Misuse of Prescription Drugs: Data from the 2002, 2003, and 2004 National Surveys on Drug Use and Health” (Colliver et al. 2006) found that after increasing for close to a decade, prescription drug abuse had leveled off between 2002-2004. This leveling off may reflect the increased awareness of and attention on this problem or perhaps the natural growth curve of this problem. Of course, the prevalence and consequence of prescription drug abuse are still very significant and its prevalence among younger persons poses significant risks for a successful transition to a healthy and productive adulthood.

The leveling off – but at relatively high levels – of prescription drug abuse is seen in Maine data. Self-reported non-medical use of prescription drugs by Maine school children has decreased from 2002 – 2006 (Table 15). (The downward trend started between 2002-2004, which is when the PMP Program was being developed, but had not been implemented.) This is very encouraging. However, in 2006, 9.4 percent of all Maine high school seniors still reported misusing prescription drugs in the past 90 days and one out of five (20 percent) had done so in the past. The dramatic increase in drug overdoses in Maine involving a pharmaceutical has also leveled off, but remains much higher than in earlier years (Table 16). The number of treatment admissions in Maine for a prescription drug abuse has continued to increase, even as admissions for other substances have declined slightly (Table 17). It is very plausible that the educational and training provided by the PMP, and the enhanced capacity to monitor and manage patients, have contributed somewhat to the leveling off of the prescription drug problem in Maine.

Table 15. Self-reported use of prescription drugs, past 30-day, by Maine school children (Mydaus) 2002, 2004, 2006

Grade	2002	2004	2006
6	3.2	2.8	1.8
7	4.5	3.7	2.0
8	7.3	6.1	3.8
9	8.8	8.9	6.2
10	10.5	11.0	8.1
11	11.3	11.6	9.5
12	10.2	10.3	9.4
Total Male	8.0	7.4	5.8
Total Female	8.0	7.9	5.6
TOTAL (ALL)	8.1	7.8	6.0

Table 16. Pharmaceutical involvement in drug related deaths in Maine 2002-2005

	2001	2002	2003	2004	2005
Total number of drug-related deaths	90	165	153	162	176
Total number of drug-related deaths for which pharmaceutical involvement is known	70	157	142	152	156

Source: Sorg M, M Greenwald, and K Marden (2007, forthcoming) Maine Drug-Related Mortality Patterns, 1997-2005. Margaret Chase Smith Policy Center, University of Maine, Orono ME.

Table 17. Number of treatment admissions by primary drug of choice

Substance	2000	2001	2002	2003	2004	2005	2006
Prescription Rx	24	23	95	159	1,690	2,099	2,469
Alcohol	6,425	6,962	7,186	7,885	8,290	7,098	6,822
Marijuana	1,338	1,553	1,653	2,046	1,900	1,569	1334
Heroin	379	519	977	1,101	1,182	1,097	1,025
TOTAL Admissions	9,750	10,971	12,479	14,332	14,946	13,283	14,019

V. Discussion

The PMP program has grown steadily since clinicians began registering for the program in January 2005, with the largest growth occurring after an on-line WEB Portal became available in March 2006. Prescribers are joining the program throughout Maine, proportionate to the distribution of the state’s population. The most common specialties among registered prescribers are family practice (304), mid-level practitioners (134), internists (90), psychiatry (55), and emergency medicine (49). These are the specialties – particularly primary care and emergency medicine – that may most benefit from the real-time availability of the data from the PMP. Prescribers have used the data from the PMP to confirm that some patients are “doctor shopping” and that others are not and are referring patients on to treatment for substance abuse, when necessary, or for further pain management. With the availability of the “real-time” information from the WEB Portal, prescribers are increasingly requesting (and using) information about new patients, who may be particularly likely to be “doctor shopping”. Prescribers are also requesting information about established patients and using this information to better manage their care. Dispensers are also using the PMP program proactively, although a bit less actively than prescribers. The confidentiality of PMP data has been maintained in an exemplary manner.

The PMP program has been successfully implemented and grown almost exactly as planned and has the wide support of stakeholders and the vast majority of participating prescribers and dispensers. The program has met / is meeting its goals to (1) give prescribers an added tool in patient care; (2) get patients who are addicted into proper treatment; and (3) ensure that those who need strong prescription drugs receive them. If the PMP program continues to grow, it appears to be on track to meet its other two goals to (4) curb the illicit use of prescription drugs in Maine, and (5) help reduce prescription drug overdoses.

Recommendations

It is likely that the PMP program will continue to grow in the near future, as newly registered prescribers and dispensers become more familiar with the WEB Portal and share their knowledge and satisfaction of the program with colleagues. There may be a leveling off of new registrants as those predisposed to join the program do so. To help promote the further growth and appropriate use of the PMP, OSA might consider the following recommendations.

Improve how up-to-date patient information is. The WEB Portal has dramatically reduced the time between when information is requested and received. However this information is usually several weeks from being totally up-to-date. There is a perception among some prescribers and dispensers that this information is more out-of-date than it is. Whatever OSA can do to improve both the timeliness, as well as the perception of the timeliness, of this information, would be helpful.

Continue to survey registered prescribers about their use of the PMP to help guide outreach and education efforts. The prescriber survey used in the implementation and in the current survey have worked well and could be sent out and analyzed by OSA staff.

Monitor and better understand which patients (age, gender, prescriptions, clinical venues, geographic area) for whom prescribers are requesting information. An analysis of requested patient history reports would be very helpful here.

Continue to provide training and outreach about the PMP program to clinicians likely to prescribe Schedule II, III, and IV drugs. The more the program is publicized, the more likely prescribing clinicians are likely to use it.

Develop a plan to better coordinate the PMP with related substance abuse and public health initiatives in the state.