Pharmacists’ experiences with dispensing opioids

Questionnaire design

The pharmacist survey was similar in content and structure to the physician survey conducted for the same project. The physician survey was based in part on questions from previous physician surveys.³ The survey asked pharmacists about the size of their practices and the communities they worked in. It asked them to estimate the number of opioid patients for whom they had concerns, the number who had been intoxicated from opioids in their pharmacies, and the prevalence of aberrant drug-related behaviour. It also asked about their views on physicians’ opioid prescribing, their interactions with physicians, and their preferred strategies for improving the safety of opioid prescribing and dispensing. The survey was pretested on a small number of community and hospital pharmacists.

Sample selection

The main inclusion criteria for the survey were an active licence and a primary practice address in Ontario (N = 10 083 in 2007), an expressed interest in research (n = 4634 of 10 083), and birth date and registration date restrictions (n = 1011 of 4634). The Ontario College of Pharmacists (OCP) mailed the surveys to the 1011 eligible pharmacists with a cover letter. A “fax-back” method was employed using a modified Dillman approach.⁴ Two reminders were mailed to the pharmacists approximately 2 weeks apart.

Analysis

Descriptive analyses of the survey results are reported. Simple measures of association including linear (Pearson R²) or rank (Spearman ρ) correlations were used to determine the relationships between certain pharmacist demographics (ie, age and years in practice) and survey items as appropriate (ie, number of total daily prescriptions in the pharmacy, number of opioid prescriptions, and number of opioid patients for whom the pharmacist had concerns). The written comments provided by respondents were collated and evaluated for common themes and suggestions to improve patient care.

Ethics

To protect the anonymity of the respondents, all mail contact with pharmacists was made by the OCP; the list of pharmacists contacted was not released to the research team by the OCP; and the survey responses were returned directly to the research team with no identifiers.

The Centre for Addiction and Mental Health Research Ethics Board approved this study. The study was funded by the Canadian Patient Safety Institute.

Description of sample

Of the 1011 surveys mailed to the pharmacists, 652 were returned, for a response rate of 64%. Of the 652 respondent pharmacists, 642 indicated that they dispensed opioid prescriptions to patients; therefore, the remainder of the analyses focused on these 642 respondents. Some or all of the 10 pharmacists who did not dispense opioids might not have worked in community settings. Pharmacists’ ages ranged from 41 to 76 years, with a mean age of 52.8 years. Years in practice ranged from 4 to 55 years, with a mean of 28.3 years. Age and years in practice were highly correlated (Pearson R² = 0.69; P < .001). Male pharmacists comprised 55% of the respondents and had a mean age of 53.5 years. Female pharmacists comprised 45% of the sample and had a mean age of 47.2 years. In contrast, the 11 900 pharmacists on the Ontario register are younger (mean age 46 years) and include a higher proportion of women (57%).

Most (77%) of the respondents practised in larger urban centres with populations of more than 30 000; 12% of respondents practised in locations with populations of 10 000 to 30 000; and 11% practised in locations with fewer than 10 000 residents. A report regarding the urban-rural distribution of Ontario pharmacists was not available from the OCP.

Opioid dispensing

Of the 642 respondents, 527 reported dispensing medications to chronic pain patients who were taking opioids daily for at least 6 months. The average estimated number of such patients attending respondents’ primary pharmacies was 50.5, with a reported range of 0 to 3200 patients. The average total number of all prescriptions (opioid and nonopioid) billed per day at the respondents’ pharmacies was 201.5 (n = 626), with a range of 3 to 12 000. The number of patients prescribed opioids for chronic pain was positively correlated with the total number of all prescriptions filled each day (Pearson R2 = 0.46; P < .001).

Concerns about opioids

Most respondents (86%) reported that they were concerned about the prescription opioid use of several (1 to 10) or many (11 or more) of their patients. The number of patients that respondents were concerned about positively correlated with the number of patients prescribed opioids for chronic pain (Spearman ρ = 0.390; P < .001). The most common opioid of concern was oxycodone (84%), followed by codeine (37%) and hydromorphone (28%).

Opioid intoxication

Of the 642 responding pharmacists, 223 (35%) reported that in the past year they had observed patients who appeared to be intoxicated by opioids while at the pharmacy (Table 1). The most common factors thought to contribute to opioid intoxication were that the patient took more than was prescribed (84%), the patient took alcohol or sedating drugs along with the opioid (70%), or the prescribed dose was too high (34%). These are the same 3 items that physicians cited as the leading causes of adverse events among their patients.³

Behaviour that could indicate opioid misuse or addiction is referred to as aberrant drug-related behaviour. The most frequently cited aberrant behaviour was requesting early renewal of the opioid prescription, with 53% reporting that this occurred sometimes and 39% reporting that it occurred frequently within the previous 3 months. Respondents indicated that they suspected that double-doctoring occurred sometimes (63%) or frequently (12%).

The third most common aberrant behaviour was requesting a replacement prescription for “lost” medications, reported as occurring sometimes by 58% and frequently by 16% of respondents (Figure 1).

• Download figure
• Open in new tab

Figure 1

Aberrant drug-related behaviour observed in the past 3 months

The survey asked pharmacists to indicate physician prescribing practices that they had observed in the previous 3 months that concerned them (Table 2). The practices noted most frequently were prescribing benzodiazepines along with opioids (20%), prescribing opioids to patients suspected of misusing them (15%), and prescribing opioids to patients who probably did not need them (13%).

Pharmacist-physician communication

Most pharmacists (69%) attempted to contact physicians about their opioid prescriptions at least once per month, with 23% attempting weekly contact. Most pharmacists reported that physicians sometimes or frequently did not respond to their concerns (61%) or were unwilling to communicate their therapeutic plans to the pharmacists (56%). Forty-three percent of pharmacists reported that physicians were frequently difficult to reach directly by telephone, while 28% said that physicians frequently did not return their calls promptly (Table 3).

Suggestions for improvement

Pharmacists were asked to rate the helpfulness of a series of strategies to improve opioid dispensing (Figure 2). The items cited most often as very or somewhat helpful were a provincial database of patients’ prescriptions (88%) and provincial prescribing guidelines (88%). Physicians also rated these 2 items as the most helpful on their survey.³

• Download figure
• Open in new tab

Figure 2

Pharmacists’ perceptions of the helpfulness of potential strategies to improve opioid dispensing: n=625.

Pharmacists’ written comments

A total of 285 respondents wrote additional comments on the survey. A common theme was that physicians should respond directly and quickly to pharmacists, and should recognize that pharmacists can help doctors improve their patient management. Another frequent theme was that physicians needed better guidance and monitoring around opioid prescribing, because their doses were too high, they prescribed to patients who either did not need them or were misusing them, and they gave early refills too readily. The respondents made numerous suggestions for improvements, such as a 1-page summary of opioid prescribing guidelines, workshops for physicians and pharmacists, a requirement that physicians write the patient diagnosis on the prescription, and a “hot-line” to identify opioid diversion and misuse.

For example, one pharmacist wrote the following: The doctors feel that we are trying to interfere with their job, they know about the patient case more than we do, they think we mostly call for authorization to refill [a prescription], and the doctor is busy .… I hope that [the provincial medical regulatory authority] lets the doctors know that they have to listen to us. When we as pharmacist[s] realize that the doctors are willing to listen to us … communication will be improved automatically.

Conclusion

Pharmacists commonly observe opioid intoxication and aberrant drug-related behaviour in their patients, and they find it difficult to communicate their concerns to physicians. These results emphasize the urgent need for a system-wide response, including national opioid prescribing guidelines, provincial prescribing databases, and strengthened protocols for physician-pharmacist communication.