Evaluating regional and temporal variations in quality indicators for opioid initiation and pharmacist medication reviews in Ontario: A population-based repeated cross-sectional study

Abstract

Background:

Quality indicators (QIs) are measures used to evaluate quality of services but are often underused in pharmacy practice. This study examines trends in 2 established QIs in community pharmacy.

Methods:

We conducted a repeated cross-sectional study in Ontario using administrative data collected between 2013 and 2023. We measured 2 QIs designed for pharmacy practice annually: (1) percentage of newly dispensed opioid prescriptions exceeding 50 morphine milligram equivalents (MME), and (2) percentage of eligible patients receiving pharmacist medication reviews within 7 days of hospital discharge. Regional differences were summarized using variance calculations, while temporal trends were analyzed using Mann-Kendall tests.

Results:

The opioid indicator demonstrated a consistent decline in the percentage of newly dispensed opioid prescriptions exceeding 50 MME across Ontario, with an absolute reduction of 10.5% from 2013 (25.6%) to 2023 (15.1%). High-dose opioid initiation ranged from 12.8% (Central) to 16.7% (West) in 2023 (range 3.9%, variance 2.3%). Significant time trends were found for all regions, with the largest reductions observed in urban regions. For the medication review indicator, provincial trends declined by 7.6%, from 16.7% in 2013 to 9.1% in 2017, followed by a modest recovery to 12.5% by 2023. Regionally, rates of medication reviews varied, with rural areas maintaining higher uptake rates compared with urban centres. Rates ranged from 7.8% (Toronto) to 16.2% (North) in 2023 (range 8.4%, variance 10.0%). A significant time trend was found only in Eastern Ontario.

Conclusion:

Significant declines in high-dose opioid initiation but inconsistent uptake of reviews across regions indicate opportunities for improvement in pharmacy practice.

Knowledge into Practice.

• Quality indicators (QIs) help to identify performance gaps and areas for improvement and can be used for audit and feedback-type interventions but are often underused in community pharmacy practice.

• High-dose opioid initiation (>50 morphine milligram equivalents) declined across Ontario, with larger reductions in urban regions.

• Pharmacist medication reviews within 7 days of discharge did not demonstrate a consistent trend, with rural areas showing higher overall uptake rates.

• Findings suggest that QI initiatives effectively reduce high-dose opioid dispensing, whereas pharmacist-led services like medication reviews face systematic barriers and uneven uptake.

Introduction

Quality indicators (QIs) are standardized measures used to evaluate the safety, effectiveness, and patient-centric approach of health care services. QIs aid in assessing performance and identifying areas for improvement in care delivery. ¹ In the context of pharmacy, QIs can offer a framework for systematic evaluation with the goal of helping pharmacists enhance medication safety, adherence, and quality of care during dispensing and clinical services. ¹ Historically, pharmacy practice has lacked systematic methods to measure the quality of care provided in both dispensing and clinical services and to establish standardized markers relative to hospital pharmacy.^1,2 This gap may be attributed to the focus of most QI development on hospital and primary care settings, where centralized data systems and integrated workflows facilitate data measurement. ² The decentralized nature of community pharmacies, combined with the absence of standardized electronic health records and established QIs, has impeded the ability to systematically improve outcomes in this setting. ² Community pharmacies provide essential services in medication management, particularly in outpatient care and posthospital discharge. Therefore, addressing this gap by introducing actionable QIs tailored for community pharmacy is necessary for pharmacy care improvement.

In Ontario, efforts led by the Ontario College of Pharmacists to bridge this gap gained traction in 2018. This collaboration led to the development of measurable and actionable QIs specifically designed for community pharmacy practice by an expert panel. ¹ These indicators focus on several measurement domains that each contain categories of quality indicators. The 2 primary domains of focus for this report are the appropriateness of dispensed medications and the transition of care domains.

Mise En Pratique Des Connaissances.

• Les indicateurs de qualité (IQ) permettent d’identifier les lacunes en matière de performance et les domaines à améliorer, et peuvent être utilisés pour des interventions de type vérification et rétroaction, mais ils sont souvent sous-utilisés dans l’exercice de la pharmacie communautaire.

• L’instauration de doses élevées d’opioïdes (équivalents à > 50 mg de morphine) a diminué dans tout l’Ontario, avec des réductions plus importantes dans les régions urbaines.

• Les examens de pharmacothérapie effectués par les pharmaciens dans les 7 jours suivant la sortie de l’hôpital n’ont pas montré de tendance constante; l’ensemble des taux d’adoption globaux étant plus élevés dans les zones rurales.

• Les résultats suggèrent que les initiatives en matière d’IQ réduisent efficacement la délivrance d’opioïdes à doses élevées, tandis que les services dirigés par des pharmaciens, tels que les examens de pharmacothérapie, se heurtent à des obstacles systématiques et à une adoption irrégulière.

The application of QI in community pharmacy remains underdeveloped compared to other health care sectors such as physicians’ clinics. ² Through analyzing data collected between 2013 and 2023, this study aims to explore 2 QIs developed for community pharmacy in Ontario: (1) the percentage of newly dispensed high-dose opioid prescriptions (>50 morphine milligram equivalents [MME]) and (2) the percentage of pharmacist medication reviews (MedsCheck) conducted within 7 days of hospital discharge. ¹ Postdischarge medication reviews have been prioritized by the Ontario Ministry of Health through the creation of a distinct MedsCheck at Discharge program (with each type of review having its own codes and requirements). This development, paired with strong evidence of the effectiveness of postdischarge medication reviews, underscores their importance in supporting safe transitions in care, particularly for patients on complex regimens like opioids. Additionally, Ontario Health Quality has included medication reconciliation at discharge as a key quality indicator, further highlighting the role of pharmacists in reducing medication-related harm and preventing readmissions.

By analyzing temporal and regional trends for these indicators, this study aims to achieve 2 objectives. The first is to demonstrate how QIs may have temporal and regional variabilities that indicate realized improvement. The second is to establish the need for continued QI measurement in the case of temporal variance. While these QIs have traditionally been established and implemented in a hospital setting, we aim to demonstrate that there is an equivalent need to measure and report on these indicators in the community setting in order to standardize the level of care provided between inpatient and community settings.

Methods

Study design

We conducted a population-based, repeated cross-sectional study to measure temporal trends and regional differences in 2 QIs for community pharmacy practice in Ontario. The indicators are reported annually and by major provincial regions (North, Central, Toronto, West, and East) in Ontario.

Data sources

We leveraged routinely collected administrative claims data held at the Institute for Clinical Evaluative Sciences (ICES). ICES is an independent, non-profit research institute whose legal status under Ontario’s health information privacy law allows it to collect and analyze health care and demographic data, without requiring consent, for health system evaluation and improvement. The use of the data for this project is authorized under section 45 of Ontario’s Personal Health Information Protection Act and did not require review by a research ethics board. All databases were linked using unique encoded identifiers and analyzed at ICES. The following ICES administrative databases were used: (1) the Narcotics Monitoring System, which captures dispensing records for all controlled substances (including opioids) dispensed from community pharmacies in Ontario, regardless of payer; (2) Ontario Drug Benefit, which captures prescription drug claims and records of pharmacist medication reviews; (3) Canadian Institute for Health Information–Discharge Abstract Database, which captures hospital discharge information; (4) Ontario Diabetes Dataset, which identifies all individuals living in Ontario with a diabetes diagnosis; (5) Registered Persons Database, which contains demographic data; (6) Drug Identification Number, which identifies opioids and chronic medications; and (7) Local Health Integration Network Database, Postal Code Conversion File, and Reference Files, which enabled regional stratification to assess geographic differences. Other databases used for applying exclusion criteria included Ontario Health Insurance Plan, National Ambulatory Care Reporting System, Continuing Care Reporting System, Resident Assessment Instrument–Contact Assessment, Resident Assessment Instrument–Home Care, and Home Care Database.

Indicator measures

We measured 2 QIs established by an expert committee prior to the initiation of this study. ¹ Both indicator definitions were aligned with provincial quality benchmarks to ensure meaningful comparisons across years and regions. ¹

The opioid dosing indicator, “high-dose opioid prescriptions”, tracks the percentage of newly dispensed opioid prescriptions exceeding 50 MME daily from community pharmacies that occurred between 2013 to 2023, enabling standardized comparisons across regions. We considered an individual to have a new dispensation if they did not have a claim for a prescription opioid for any indication in the 6 months prior to the index dispense date. Only opioid prescriptions intended for analgesia and given in oral or transdermal forms were included. Opioids prescribed for cancer, opioid agonist therapy, cough suppressants, antidiarrheals, and opioids in injectable forms were not captured by our data. The discharge medication review indicator measures the percentage of eligible patients who received pharmacist medication reviews within 7 days of discharge (including the discharge date). We included urgent acute inpatient hospitalizations between 2013 and 2023 where patients were discharged alive to their home and presented to the pharmacy (flagged with a dispensation) within 7 days. These indicators were defined based on alignment with provincial definitions; more details can be found in existing documentation.^1,3

Statistical analysis

We evaluated trends over time and regional differences for the QIs of interest from 2013 to 2023. That is, annual trends were reported for Ontario as a whole and stratified by region, with data being visualized using line graphs. For both indicators, we stratified by region in order to identify those with persistently high rates of high-dose opioid initiation for the first indicator and differences in uptake of pharmacist medication reviews for the second indicator. Regional trends were overlaid to visually identify areas with the greatest improvements and those with persistent underperformance.

Temporal variation, or overall trend over time, was measured using the Mann-Kendall test to assess the presence of monotonic trends over time. Regional variation, or differences between regions, were measured using 2 metrics: variance and range. Variance was calculated through a sample variance equation, $S^2=[\Sigma {(xi-\overline{x})}^2]/n-1$ , where xi is the regional percentage, $\overline{x}$ is the mean percentage, and n is the number of regions. High variance may suggest inequity or lack of standardization in care delivery, pinpointing areas for improvement. Variance was calculated annually to quantify deviations in performance relative to regional averages over time. An annual variance that is higher than a previous year would indicate that provincial regions have larger differences with respect to that quality indicator. Conversely, decreasing regional variance over time may suggest better consistency in performance across regions. Range was also reported as a proxy for regional differences over time, determined by calculating the difference between the maximum and minimum percentages for the displayed years. This statistical approach enabled the identification of significant differences in performance across regions.

Results

We observed a consistent decline in the percentage of newly dispensed opioid prescriptions exceeding 50 MME across Ontario from 2013 to 2023. Figure 1A illustrates this temporal trend at the provincial level, showing an absolute reduction of 10.5%, from 25.6% in 2013 to 15.1% in 2023. Regional trends, highlighted in Figure 1B, revealed notable differences in prescribing patterns geographically. The largest reductions were observed in urban regions such as Toronto (13.5% absolute decline, from 28.0% to 14.5%) and Western Ontario (9.3% absolute decline, from 26.0% to 16.7%). Conversely, Northern Ontario showed the slowest decline, with an absolute reduction of 7.0% (from 23.6% to 16.6%). High-dose opioid initiation ranged from 12.8% (Central) to 16.7% (West) in the final year of study, with a range of 3.9% and a variance of 2.3% (Table 1). The Mann-Kendall trend analysis indicated that all regions had statistically significant trends over time in the opioid indicator (Table 2).

Figure 1.

Percentage of unique individuals newly dispensed an opioid used to treat pain with a high daily dose (>50 morphine milligram equivalents) between 2013 and 2023: (A) provincial and (B) regional

Table 1.

Regional variance in quality indicators: High-dose opioid prescriptions and discharge pharmacist medication reviews (2013–2023)

Indicator	Overall range for province, 2013–2023	Lowest region in 2013	Highest region in 2013	Lowest region in 2023	Highest region in 2023	Range in 2013*	Variance in 2013 (calculated)	Range in 2023*	Variance in 2023 (calculated)
Opioids	15.1–25.6	23.4 (Central)	28.0 (Toronto)	12.8 (Central)	16.7 (West)	4.6	3.2	3.9	2.3
Transitions of care	9.1–17.2	9.9 (Toronto)	18.4 (North)	7.8 (Toronto)	16.2 (North)	8.5	9.8	8.4	10.0

Note: All values are expressed as percentages.

^*Determined as maximum value minus minimum value.

Table 2.

P-values from Mann-Kendall trend analysis for opioid prescriptions and pharmacist medication reviews by region (2013–2023)

Region	Opioid indicator p-values	Medication review indicator p-values
Overall	<.001	0.276
Central	<.001	0.350
East	<.001	0.020
North	<.001	0.350
West	<.001	0.276

Note: p < 0.05 is statistically significant, indicating a trend in the data not due to random variation.

We observed notable variability in the prevalence of pharmacist medication reviews after hospital discharge across Ontario. Provincial trends, depicted in Figure 2A, showed an initial 7.6% absolute decline in medication review rates from 16.7% in 2013 to 9.1% in 2017, followed by a modest recovery to 12.5% by 2023. Regional analyses, as summarized in Figure 2B, highlighted notable differences. Medication review rates postdischarge were highest in rural areas such as Northern Ontario, although they fell by 1.7% over the study period, from 18.4% in 2013 to 16.7% in 2023. In contrast, urban centres such as Toronto had consistently lower rates of pharmacist medication reviews postdischarge, which fell 2.1%, from 9.9% to 7.8% over the same period. Rates ranged from 7.8% (Toronto) to 16.2% (North) in the final year of study, with a wider range of 8.4% and a variance of 10.0% (Table 1). The trend analysis indicated that a statistically significant time trend was found for Eastern Ontario only (Table 2).

Figure 2.

Percentage of eligible individuals who had a medication review within 7 days of being discharged home from hospital between 2013 and 2023: (A) provincial and (B) regional

Appendix Figures A and B, available online under Supplementary Materials, provide additional granularity on subregional differences in both indicators.

Discussion

This study explored regional and temporal trends among 2 different QI indicators used in community pharmacy practice in Ontario. We found that high-dose opioid initiation consistently declined, with some variation across regions, whereas pharmacist medication reviews after hospital discharge showed much more variation geographically and over time. These findings demonstrate the significant opportunity for improvement across the province in core indicators of high-quality care and the potential role of pharmacists in driving these improvements.

The first indicator showed a consistent reduction in high-dose opioid initiation, with a consistent decline provincially of over 10% in the past decade. However, rates of decline varied across the province, with some regions showing slower progress. Given that the prescribing authority for opioids rests with the physician, most QI efforts related to opioid prescribing have targeted physicians. For example, Canadian clinical practice guidelines strongly recommend restricting prescribed opioid doses to less than 90 MME daily among patients with chronic non-cancer pain. ⁴ Other provincial and national initiatives have also aimed to educate physicians on safe and appropriate opioid prescribing.^5,6 Our results suggest that these efforts have likely been beneficial and effective in reducing high-dose opioid prescribing. However, pharmacists may have a contributing role in sustaining these improvements, such as flagging high-dose opioid prescriptions when they are received at the pharmacy and communicating directly with prescribers to discuss changing dose. We note that the intent of this indicator is not to achieve a value of zero but rather to facilitate benchmarking against peers and to explore patterns of variation.

In contrast, pharmacists have more capacity to directly affect the performance of the second indicator. We found larger temporal and regional differences for the medication review indicator, which may be reflective of less measuring and reporting of this indicator directly to pharmacies and highlights a gap in the implementation of pharmacist medication reviews postdischarge. The marked regional variation emphasizes the necessity of region-specific interventions to enhance service uptake, particularly in consistently underperforming areas. Barriers to medication review uptake are influenced by several operational challenges that community pharmacies face, such as high work volume, limited staffing, and workflow inefficiencies.^7,8 Pharmacists often encounter significant time constraints with competing priorities, making it difficult to allocate adequate time for comprehensive medication reviews.^7,8 Additionally, limited staffing and high patient volumes further exacerbate the challenge, leaving pharmacists with insufficient capacity to provide timely medication review services.^7,8 Workflow inefficiencies and lack of integration with other health care providers also contribute to the inconsistent delivery of the program.^7,8 Addressing these barriers while using QI to identify which community pharmacies require aid may improve medication review uptake and reduce regional differences. The medication review indicator also exhibited greater temporal inconsistency, with initial declines in uptake followed by partial recovery in later years. This variation shows the necessity of continuous measurement to track performance, identify deviations, and implement interventions. The significant decline in pharmacist medication reviews performed from 2016 to 2017 may be attributed to the change in policy surrounding MedsCheck services that resulted in increased documentation requirements and added mandatory components to the service. ⁹ The levels never returned to prepolicy values, suggesting that while administrative complexity may enhance standardization and quality in pharmacy services, it also leads to barriers in service delivery. ⁹ Mitigating this impact requires implementation of streamlined documentation processes and digital automation tools, and flexible service models have been proposed to reduce administrative burden while maintaining quality standards. ⁹ Additionally, incentivizing pharmacist participation and integrating medication reviews into routine pharmacy workflows through targeted outreach and patient engagement could improve accessibility. These strategies will help ensure that quality improvement efforts do not inadvertently hinder service uptake and promote more consistent and standardized access to these services provincially. ⁹

We recognize that the COVID-19 pandemic beginning in 2020 may have affected the performance of these QIs. An Ontario study found that the number of new opioid users declined after the lockdown, ¹⁰ possibly due to reduced health care access and fewer in-person consultations, causing physicians to be more cautious when initiating opioid therapy. Therefore, the pandemic may have contributed to the continuing decline found for the high-dose opioid indicator. The role of pharmacists also rapidly expanded during the pandemic as a response to health system changes, with pharmacists bearing greater responsibilities such as managing drug shortages, providing patient education, and delivering influenza and COVID-19 vaccinations.^11,12 This in turn may have reduced pharmacists’ capacity to conduct medication reviews, which could explain the sudden drop in medication reviews in 2020 following a continuous rise in the previous few years.

This study underscores the value of QIs in measuring pharmacy service outcomes and guiding improvements in clinical practice. In summary, the significant downward trend in the opioid prescribing indicator suggests that targeted interventions at the physician and pharmacy level have been effective, while still having regional differences, whereas the lack of a trend in the pharmacist medication review indicator indicates that barriers to its implementation may be more complex and require nuanced solutions to overcome the barriers, resulting in temporal and regional differences. The differences reflect the variance among regional capacities to implement and sustain improvement strategies effectively and reiterate the need for QI implementation in community pharmacy. To amplify these improvements, implementing QI reporting at the pharmacy level could yield actionable insights, enabling individual pharmacies to benchmark their performance, identify specific deficiencies, and adopt targeted strategies for quality enhancement. ¹³ Continued monitoring and localized reporting present critical opportunities to sustain progress and address evolving challenges in pharmacy practice. Without ongoing monitoring, it becomes difficult to assess whether improvements are sustained. ¹⁴ The findings indicate that periodic evaluation of QIs is required to inform evidence-based decision-making and support adaptive quality improvement efforts over time. ¹⁴ Our findings align with some current literature that identified that continuous quality improvement (CQI), which is a progressive incremental improvement process measured regularly, reduces QI variability in health care settings. ¹⁴

Identifying the need for QI in community pharmacy care is important; however, implementation can be a challenge due to actionability. Creating actionable QIs in community pharmacy ensures that they translate into meaningful improvements in practice. Actionable QIs are those that can be directly influenced by pharmacy interventions and have a clear link to patient outcomes. For example, the medication review indicator tracks the percentage of eligible patients who receive a pharmacist medication review within a defined timeframe. This QI is actionable because pharmacies can implement targeted strategies, such as automated reminders or delegation of tasks to pharmacy technicians, to improve uptake rates. Similarly, high-dose opioid prescribing is a measurable and actionable QI, as it allows pharmacies to assess prescribing patterns and intervene by offering alternatives to high-dose prescriptions. Actionable QIs improve engagement among health care providers by offering clear performance targets and fostering accountability. In community pharmacy, where operational environments are diverse, ensuring that QIs are actionable helps overcome fragmentation and supports improvement of care.

QIs themselves require periodic reassessment to maintain their relevance as health care systems, patient populations, and regulatory requirements adapt to changes in the health care landscape. QIs may initially drive improvement, but a lack of continued measurement can result in a gradual decline in care quality. ¹⁵ Furthermore, as health care practices change over time, QIs may become less appropriate, reducing their ability to effectively measure performance or inform practice. ¹⁵ This further necessitates a need to consistently measure and update QIs to ensure they continue to reflect current health care priorities and maintain their use in quality improvement efforts.

QIs are predominantly used in hospital settings and have been created by the Canadian Society of Hospital Pharmacists to drive improvement in hospital pharmacy services. ¹⁶ Hospital systems have centralized systems that facilitate data collection. In contrast, community pharmacies lack integrated systems and many lack standardized processes when faced with large work volumes, which complicates the development and application of QIs.^16,17 Community pharmacies are crucial in medication management for patients who are not seen in inpatient settings, and the absence of standardized QIs in this setting limits the ability to evaluate performance and implement systematic improvements. The findings of this study suggest that the introduction of QIs in community pharmacy practice is necessary to enhance accountability, improve service delivery, and ensure consistency in care. Implementing QIs such as the opioid and medication review indicators can provide a standardized approach to evaluating pharmacy services and support broader health care system goals related to improving patient safety and providing continuity of care from hospital to community settings.

Limitations

This study is subject to some limitations. The administrative data used are limited by quality and input accuracy of the data, which are recorded by the providers. While the study examined temporal trends and regional variability, it did not account for potential confounding factors such as changes in health care policies, socioeconomic status of patients, or pharmacy-level operational differences. Furthermore, our administrative data could not capture the appropriateness of opioid prescriptions; therefore, we are unable to ascertain whether the high-dose prescriptions were indicated, nor could we identify patients who could have benefited from receiving high-dose opioids but did not, possibly leading to undertreatment of pain. The study was limited to Ontario, which may reduce the generalizability of findings to other provinces or countries with different health care systems and pharmacy practice models. Last, this study only measured quantitative data, which do not capture qualitative feedback from pharmacy staff that may have provided additional context for barriers and facilitators of QI implementation and factors that ultimately could affect variance.

Conclusion

This study highlights the importance of QI indicators in assessing and improving pharmacy services such as high-dose opioid dispensing and postdischarge pharmacist medication reviews. The findings on regional variance support the importance of implementing quality indicators in community pharmacy services and emphasize the need for sustained efforts in monitoring QI performance across different regions to ensure consistent and equitable care. Furthermore, the variation in temporal trends highlights the importance of continuous measurement. The statistically significant decline in high-dose opioid initiation reflects the potential for physician-targeted QI initiatives to guide meaningful improvements, while the inconsistent uptake of medication reviews is an example of a pharmacist-led service that has many systematic and operational barriers to achieve equitable outcomes. Given that pharmacists have more direct influence in providing medication reviews compared to opioid prescribing, there exists opportunities for pharmacists to play a role in promoting high-quality care and for the potential collaboration between physicians and pharmacists to maximize the impact of these QI indicators. Findings from this analysis will inform future strategies for improving QI measurement processes. Future research should focus on updating, improving, and augmenting QIs in community pharmacy settings within the expanded scope of practice among pharmacists. ■