Concept: Primary Care Quality Indicators

Introduction

This concept identifies and describes 29 primary care quality indicators that are used to measure and evaluate the differences in primary care service delivery provided in the Winnipeg Health Region. The primary care quality indicators fall into four broad categories: prevention and screening, chronic disease management, medical care, and health services use and delivery. The concept provides information / links to detailed technical information that describes how each of these indicators are developed using data in the MCHP Data Repository.

Information for this concept is taken directly from the MCHP deliverable A Comparison of Models of Primary Care Delivery in Winnipeg by Katz et al. (2016), which built off of previous research investigating primary care quality indicators and the Physician Integrated Network (PIN) (Katz et al. 2010 and 2014).

This concept also provides additional information on the relationship between primary care quality indicators and models of primary care delivery and social complexities / social complexity index that were investigated in this research. Links are provided to the relevant sections of the deliverable for access to more information and discussion.

Additional information in this concept provides a summary description of the previous research investigating primary care quality indicators and family physicians and the Physician Integrated Network (PIN) by Katz et al. (2004, 2010 and 2014). The research into primary care quality indicators began in 2004 with 13 indicators and has evolved into the 29 indicators presented in the 2016 research. Links are provided to specific definitions, discussions and results of this previous research.

Primary Care Quality Indicators Defined

The following is a list of the 29 primary care quality indicators that were investigated and reported on in Katz et al. (2016). These indicators are divided into four categories:

I. - prevention and screening,
II. - chronic disease management,
III. - medical care, and
IV. - health services use and delivery.

For each indicator listed, a brief definition is provided as well as a link to the technical details documented in Appendix 1 - Primary Care Quality Indicator Definitions of the deliverable. The technical details include: a definition, data source(s), the tariff / ICD / ATC code values that are relevant, and any conditions, restrictions, and exclusions that are applied. A link is also provided to the relevant research methods / definitions documented in the on-line MCHP Concept Dictionary and Glossary.

I. Prevention and Screening

This category includes seven different primary care quality indicators, including:

1. breast cancer screening for women aged 50-74 - this indicator is defined as the proportion of women aged 50-74 who received at least one mammogram in a two-year period (fiscal years 2011/2012 to 2012/2013). Women with a history of breast cancer were excluded from this analysis because this indicator was designed to measure a "preventive" procedure.
   
   • Technical Definition - Appendix 1
     
   • Breast Cancer Screening / Mammography concept.
   
   
2. cervical cancer screening for women aged 21-69 - this indicator is defined as the proportion of women aged 21-69 who had a least one Papanicolaou (Pap) test in a 3-year period (fiscal years 2010/11-2012/13). Women who had a hysterectomy were excluded from the analysis.
   
   • Technical Definition - Appendix 1
     
   • Cervical Cancer Screening (Pap Test) concept.
   
   
3. colorectal cancer screening for adults aged 50-74 - this indicator is defined as the proportion of male and female patients aged 50-74 who had at least one Fecal Occult Blood Test (FOBT) in a two-year period (fiscal years 2011/12-2012/13).
   
   • Technical Definition - Appendix 1
     
   • Colorectal Cancer Screening / Fecal Occult Blood Test (FOBT) concept.
   
   
4. completed vaccinations at age two - this indicator is defined as the percent of two-year old children with health insurance coverage in Manitoba from birth to their 2nd birthday, who received all of the vaccinations recommended in the provincial vaccination schedule. The Manitoba vaccination schedule changed during the study period, and this change was taken into consideration in the analysis.
   
   • Technical Definition - Appendix 1
     
   • Childhood Immunizations concept.
   
   
5. annual influenza vaccination, adults aged 65 and older - this indicator is defined as the proportion of patients aged 65 and older who received the influenza vaccine in each year of the study period in which they were eligible for inclusion in the indicator age span.
   
   • Technical Definition - Appendix 1
     
   • Adult Immunizations concept.
   
   
6. annual influenza vaccination, people with total respiratory morbidity (TRM) - this indicator is defined as the proportion of patients diagnosed with TRM (acute bronchitis, chronic bronchitis, bronchitis not specified as acute or chronic, emphysema, asthma, and chronic airway obstruction) who received the influenza vaccination in each year of the study period. TRM was defined as at least one hospitalization in one year or at least one primary care provider visit in one year with a diagnosis of one of the illnesses included in TRM.
   
   • Technical Definition - Appendix 1
     
   • Adult Immunizations concept.
   
   
7. pneumococcal vaccination, adults aged 65 and older - this indicator is defined as the proportion of patients aged 65 and older at the beginning of the study period (April 1, 2010) who had ever received a pneumococcal vaccination. This vaccination was introduced in Manitoba in 2001.
   
   • Technical Definition - Appendix 1
     
   • Adult Immunizations concept.

II. Chronic Disease Management

This category includes eight different primary care quality indicators, including:

1. diabetes management: eye examination - this indicator is defined as the percentage of patients aged 20-79 who had one eye examination by an ophthalmologist or optometrist for each year that they were diagnosed with diabetes.
   
   • Technical Definition - Appendix 1
     
   • Diabetes - Measuring Prevalence
     
   • Diabetes Care: Prevalence of Annual Eye Exams glossary term.
   
   
2. congestive heart failure (CHF) management: initiation of ACE inhibitor drug treatment - this indicator is defined as the proportion of newly diagnosed CHF patients (diagnosed for the first time during the study period) aged 20 and older who filled a prescription for either angiotensin-converting enzyme (ACE) inhibitors or angiotensin II receptor blocker (ARB) within 3 months of diagnosis.
   
   • Technical Definition - Appendix 1
     
   • Congestive Heart Failure (CHF) concept.
   
   
3. congestive heart failure (CHF) management: persistence of ACE inhibitor drug treatment - this indicator is defined as the proportion of CHF patients aged 20 and older who were persistent users of either ACEI or ARB, as defined by their prescriptions being filled for 80% of the days between CHF diagnosis and the end of study period.
   
   • Technical Definition - Appendix 1
     
   • Congestive Heart Failure (CHF) concept.
   
   
4. post-myocardial infarction (MI) management: initiation of beta-blocker drug treatment - this indicator examines the odds ratio of newly diagnosed MI patients aged 20 and older who filled at least one beta-blocker prescription within 4 months of hospital discharge at the time of the MI.
   
   • Technical Definition - Appendix 1
     
   • Post MI Management with Beta-Blockers section of the Acute Myocardial Infarction (AMI) / Myocardial Infarction (MI) - Definition, Rates and Post-AMI Management concept.
   
   
5. post-myocardial infarction (MI) management: persistence of beta-blocker drug treatment - this indicator examines the odds ratio of MI patients aged 20 and older who were persistent users of beta-blockers, as defined by having filled prescriptions for 80% of the days between MI diagnosis and the end of the study period.
   
   • Technical Definition - Appendix 1
     
   • Post MI Management with Beta-Blockers section of the Acute Myocardial Infarction (AMI) / Myocardial Infarction (MI) - Definition, Rates and Post-AMI Management concept.
   
   
6. post-myocardial infarction (MI) management: initiation of cholesterol-lowering drug treatment - this indicator examines the odds ratio of newly diagnosed MI patients aged 20 and older who filled at least one cholesterol-lowering prescription within 4 months of hospital discharge following the MI.
   
   • Technical Definition - Appendix 1
     
   • Post MI Management with Cholesterol Lowering Drug Treatment section of the Acute Myocardial Infarction (AMI) / Myocardial Infarction (MI) - Definition, Rates and Post-AMI Management concept.
   
   
7. post-myocardial infarction (MI) management: persistence of cholesterol-lowering drug treatment - this indicator examines the odds ratio of MI patients aged 20 and older who were persistent users of cholesterol lowering drugs, as defined by having filled prescriptions for 80% of the days between MI diagnosis and the end of the study period.
   
   • Technical Definition - Appendix 1
     
   • Post MI Management with Cholesterol Lowering Drug Treatment section of the Acute Myocardial Infarction (AMI) / Myocardial Infarction (MI) - Definition, Rates and Post-AMI Management concept.
   
   
8. asthma management: medication use - this indicator is defined as the proportion of patients aged 20 and older with two or more prescriptions for beta2-agonists (relievers) within 12 months of the study period who filled a prescription for a medication recommended for long-term control of asthma, such as inhaled corticosteroids and leukotriene antagonists. The analysis excluded patients with COPD, which was defined as having filled one or more prescriptions of ipratropium bromide.
   
   • Technical Definition - Appendix 1
     
   • Asthma Care Controlled by Medication Use concept.

III. Medical Care

This category includes three different primary care quality indicators, including:

1. benzodiazepine prescribing in community dwelling adults aged 75 and older - this indicator is defined as the percentage of patients aged 75 and older who had at least two prescriptions for benzodiazepines or at least one prescription for benzodiazepines with a greater than 30-day supply, measured annually for three fiscal years.
   
   • Technical Definition - Appendix 1
     
   • Benzodiazepine Use concept.
   
   
2. Beers drug prescribing in community dwelling adults aged 65 and older - this indicator is defined as the percentage of community-dwelling patients aged 65 and older who filled at least one prescription for a drug which the Beers Criteria suggest should be avoided. The list of drugs used in this study can be found in Appendix 1.
   
   • Technical Definition - Appendix 1
     
   • Beers Criteria glossary term.
   
   
3. depression care, prescription follow-up - this indicator is defined as the percentage of patients diagnosed with depression who filled a prescription for an antidepressant medication within two weeks of the diagnosis and made three subsequent ambulatory visits within four months of the prescription being filled (any diagnosis, any primary care provider). Patients with a prescription for antidepressants or a diagnosis of depression within the two years prior to the first diagnosis in the study period were excluded from this analysis. A higher rate of this indicator is considered a better outcome.
   
   • Technical Definition - Appendix 1
     
   • Depression - Defining in Administrative Data concept.
     
   • Antidepressant Prescription Follow-Up glossary term.

IV. Health Services Use and Delivery

This category includes eleven different primary care quality indicators, including:

1. continuity of care of assigned primary care provider - this indicator is defined as the extent to which an individual patient sees their assigned primary care provider over a specified period of time. This indicator uses an index, called the Continuity of Care Index (COCI), that weights both the frequency of ambulatory visits to each primary care provider and the dispersion of ambulatory visits among primary care providers.
   
   • Technical Definition - Appendix 1
     
   • Measuring Continuity of Care concept.
     
   • Continuity of Care Index (COCI) - Physician glossary term.
   
   
2. clinic-based continuity of care - this indicator is defined as the extent to which an individual patient sees primary care providers in the same clinic compared to providers in other clinics. This indicator also uses the Continuity of Care Index (COCI), and weights both the frequency and dispersion of ambulatory visits to different clinics.
   
   • Technical Definition - Appendix 1
     
   • Measuring Continuity of Care concept.
     
   • Continuity of Care Index (COCI) - Physician glossary term.
   
   
3. referral rates - this indicator is defined as the average number of referrals per patient. All referrals (i.e., referrals made by assigned and unassigned providers) to specialists were counted. Referrals to pathology, radiology, other primary care providers and to obstetricians for uncomplicated pregnancy are excluded.
   
   • Technical Definition - Appendix 1
     
   • Referral Rates for Primary Care Physicians to Specialists concept.
   
   
4. hospitalizations for ambulatory care sensitive conditions (ACSC): adults aged 74 and younger - this indicator is defined as the rate at which patients aged 0-74 were hospitalized for a select group of the ACSCs, per 1000 patients. Only four ACSCs were considered, including: chronic obstructive pulmonary disease (COPD), asthma, congestive heart failure (CHF) and diabetes.
   
   • Technical Definition - Appendix 1
     
   • Ambulatory Care Sensitive (ACS) Conditions concept.
   
   
5. x-ray for lower back pain: adults aged 20 and older - this indicator is defined as the percentage of patient 20 years and older who had at least 1 lower back pain X-ray (at least one of the following tariff codes: 7034, 7035, 7036, 7037, 7038, 7039, 7041, 7054, 7061, 7193) and a lower back pain diagnosis (ICD-9-CM code: 724) over the 3 year study period (fiscal years 2010/11 – 2012/13).
   
   • Technical Definition - Appendix 1
   
   
6. hospital episodes with a readmission within 30 days - this indicator is defined as the percentage of unplanned, inpatient readmissions to an acute care facility (the same or different hospital) within 30 days following discharge from the original hospital episode.
   
   • Technical Definition - Appendix 1
     
   • Readmissions to Hospital concept.
   
   
7. emergency department visit rate for patients with CTAS 4 or 5 - this indicator is defined as the proportion of patients coming to an emergency department (ED) during the three-year study period for conditions that are defined as "less urgent" (CTAS 4) or "non-urgent" (CTAS 5); that is, this indicator measures the potentially inappropriate use of EDs. This indicator assumes that inappropriate ED use is a reflection of poor primary care access or dissatisfaction of the primary care received by the patient. A lower rate of this indicator is considered a better outcome.
   
   • Technical Definition - Appendix 1
     
   • Canadian Triage & Acuity Scale (CTAS) - Emergency Department glossary term.
   
   
8. ambulatory visits to primary care - this indicator is defined as the number of visits per patient to a licensed primary care provider in an outpatient setting in Winnipeg, annualized over the three-year study period. All visits the patients made to primary care providers were counted, regardless of the model of primary care they received or participated in.
   
   • Technical Definition - Appendix 1
     
   • Ambulatory Visits - Physician concept.
     
   • Models of Primary Care Delivery concept.
   
   
9. ambulatory visits to primary care for patients with Resource Utilization Bands (RUB) 3, 4 or 5 - this indicator is defined as the number of visits per patient to a licensed primary care provider in an outpatient setting in Winnipeg, annualized over the three-year study period. Only patients with a Resource Utilization Bands (RUB) of "moderate" (3), "high" (4), or "very high" (5) morbidity are included in the analysis.
   
   • Technical Definition - Appendix 1
     
   • Ambulatory Visits - Physician concept.
     
   • Resource Utilizations Bands (RUB) section of the Adjusted Clinical Groups® (ACG®) - Overview concept.
   
   
10. ambulatory visits to primary care for patients with three or more social complexities - this indicator is defined as the number of visits per patient to a licensed primary care provider in an outpatient setting in Winnipeg, annualized over the three-year study period. Only patients with three or more social complexities are considered in the analysis.
    
    • Technical Definition - Appendix 1
      
    • Ambulatory Visits - Physician concept.
      
    • Social Complexities / Social Complexity Index concept.
    
    
11. ambulatory visit with a call to Health Links – Info Santé within two days of visit - this indicator is defined as the rate of ambulatory visits to primary care over the study period with at least one associated call made to Health Links - Info Santé within a 48-hour period after the ambulatory visit. The rate presented is per 1,000 ambulatory visits. A lower rate of this indicator is considered a better outcome.
    
    • Technical Definition - Appendix 1
      
    • Ambulatory Visits - Physician concept.
      
    • Health Links - Info Santé glossary term.

Analyses of Primary Care Quality Indicators by Model of Primary Care and by Social Complexity

In Katz et al. (2016), the main objectives were to compare and evaluate primary care quality indicators for the five models of primary care delivery in Winnipeg, and to describe the impact of social complexities on primary care quality indicators. The results of this investigation reported on each primary care quality indicator, and looked at the eligible population and crude rates, a comparison between models of primary care, the relationship between social complexities, and the relationships between the number of social complexities.

Statistical Analyses Methods

In this research, comparisons between models of primary care were made using generalized linear mixed models and calculated at a type 1 error rate (i.e., alpha) of 0.01 due to the multiple comparisons being performed. These statistical models allow for comparisons among the results for the models of primary care for each primary care quality indicator, while accounting for clustering of patients within primary care providers and clinics. Odds ratios and relative rates were calculated using logistic and Poisson regression, respectively. The Continuity of Care Index (COCI) is a continuous outcome; therefore, a multilevel linear regression was used for the comparisons between models of primary care. Statistical model results for each of the primary care quality indicators are presented in Online Appendix 3 - Model Results of the deliverable.

Results and Discussion

The results of this research on primary care quality indicators are presented and discussed within specific chapters of the deliverable. Direct links are provided below to the beginning of each chapter, corresponding to the four broad categories of primary care quality indicators. For each primary care quality indicator, tables and graphs are presented. For most indicators, this includes:


• a table identifying the eligible population and crude rates by model of primary care;
  
• a table comparing the "basic" adjusted indicator results among the models of primary care, adjusting for patient characteristics (i.e., age, sex, income quintile, and resource utilization band [RUB]) and primary care provider characteristics (i.e., age, sex, years in practice, and country of graduation);
  
• comparison of the "full" adjusted indicator results among the models of primary care, adjusting for social complexities, in addition to patient and primary care provider characteristics;
  
• a table describing the relationship between each social complexity and the indicator outcome; and
  
• a figure illustrating the relationship between the number of social complexities (compared to zero social complexities) and the indicator outcome, as well as a linear trend test; when the linear trend test result is significant, this means that the indicator result either decreases or increases as the number of social complexities increases.

Only statistically significant results are described in the text and shown in the tables. In some cases, where no or very few statistically significant results occurred, tables or graphs are not shown and this is indicated within the text.

The findings describe the differences in quality of care among the five models of primary care delivery. However, the non-PIN FFS model is sometimes used as a reference model against which all other models are compared, because it is the dominant model of primary care in Winnipeg. For many of the indicators, social complexity is associated with poorer quality of care results across the models of care.

The details of the results and discussion for each primary care quality indicator are available in the following chapters of the deliverable:


• Chapter 4: Prevention and Screening Indicator Results;
  
• Chapter 5: Chronic Disease Management Indicator Results;
  
• Chapter 6: Medical Care Indicator Results; and
  
• Chapter 7: Health Services Delivery and Use Indicator Results

Additional detailed results from the investigation of primary care quality indicators in Katz et al. (2016) are available in two on-line appendices:

• On-line Appendix 2 - Primary Care Quality Indicator Crude Rates by Clinic presents the population and crude rate (%) for each of the 29 primary care quality indicators investigated in this research for each of the individual clinics in each model of primary care (excluding Non-PIN FFS clinics).
  
  
• On-line Appendix 3 - Model Results presents the statistical model output (odds ratio (95% confidence limits) and p-values) for each of the 29 primary care quality indicators investigated in this research for different effects, including model of primary care comparisons, demographic characteristics (e.g.: patient age, Resource Utilization Bands (RUBs), income quintiles, patient sex (male vs. female), years of provider practice) and social complexities.

Previous Research Using Primary Care Quality Indicators

Three previous MCHP research projects investigated primary care quality indicators in relation to family physicians and the Physician Integrated Network (PIN) project. The projects' goals and use of primary care quality indicators are summarized below, and direct links are also provided to lists and definitions of the indicators used in each project, as well as to the results of this research.

Katz et al. (2004)

In the Using Administrative Data to Develop Indicators of Quality in Family Practice deliverable by Katz et al. (2004), they developed and investigated 13 quality of care indicators that measured family physician services / behaviors. These indicators were divided into two groups:


1. Disease Prevention / Health Promotion Indicators - includes five measures: childhood immunization, influenza vaccination, cervical cancer screening, cholesterol screening, and blood sugar screening.
   
2. Acute and Chronic Disease Management Indicators - includes eight measures: antidepressant prescription follow-up, asthma care, potentially inappropriate prescribing of benzodiazepines for older adults, diabetes care: cholesterol testing, diabetes care: eye examination, post-myocardial infarction (MI) care: Beta-Blocker prescribing, and post-myocardial infarction (MI) care: cholesterol testing.

For a list of codes used to define these quality of care indicators, please see Table 4 in the deliverable.

A discussion and presentation of the results for each of the indicators is available in Sections 3.1, 3.2 and 3.3 of the deliverable.

Katz et al. (2010)

In the Physician Integrated Network Baseline Evaluation: Linking Electronic Medical Records and Administrative Data deliverable by Katz et al. (2010), they compared the electronic medical record (EMR) and administrative data for data fields (variables) that are common to both data sources, and identified the indicators that are feasible to measure using administrative data from the MCHP Data Repository. This initial study involved four PIN clinics.

In total, 15 indicators of preventive care and chronic disease management were investigated in this research. The definitions for these indicators initially followed the Canadian Institutes for Health Information (CIHI) Primary Care Indicators definitions - see CIHI Website - Primary Health Care for more information. In MCHP research, some of these indicators were changed as they were implemented, due to data limitations. All 15 indicators were measured using data from the MCHP Data Repository. In addition, each individual PIN clinic focused only on a small group of indicators for Phase 1 of PIN, and therefore, the majority of indicators were not available from all clinics.

The indicators of quality primary care that were analyzed for this project, along with definitions for each indicator, are available in Table 3.4: Indicators of Quality Primary Care of the deliverable. Appendix Table A1.1: Codes Used to Define the Indicators lists the corresponding tariff, ICD diagnoses, and/or medication (ATC) codes for each indicator as well as other conditions / restrictions that are applied.

Results for each of the indicators are available in the section titled How the Indicator Results are Presented. This includes a short definition, the eligible population (from the EMR, from the MCHP Data Repository and from both data sources) and the outcome rates. The primary interest in this research is not the comparison between clinics, but rather the within-clinic comparison between the expected rate for that clinic and the different clinic populations (EMR and MCHP only).

Katz et al. (2014)

In the Physician Integrated Network: A Second Look deliverable by Katz et al. (2014), the goal of this study was to measure the impact of the PIN initiative on the quality of care delivered by PIN clinics. This study involved 12 PIN clinics.

A total of 23 primary care quality indicators were investigated, divided into three categories; prevention and screening, disease management, and healthcare delivery. A list of the primary care quality indicators analyzed in this project are available in Table 2.4: Study Indicators and Quality Based Incentive Funding (QBIF) Indicators of the deliverable. This table identifies those indicators associated with quality based incentive funding (QBIF) that can be measured using MCHP Data Repository data. Appendix Table 1.1: Table of Definitions and Codes of Indicators lists the corresponding tariff, ICD diagnoses, and/or medications (ATC) codes for each indicator, as well as other conditions / restrictions that are applied. Appendix Table 1.2: Quality Based Incentive Funding (QBIF) Indicators lists each of the PIN QBIF indicators, and includes a definition for each indicator.

Results of this study are presented in different chapters of the deliverable. Chapters 3-5 present outcomes for primary care quality indicators in each of the three categories. Chapter 6 presents information on pre- and post-implementation and long term follow-up outcomes for four of the PIN clinics. Chapter 7 presents indicator outcomes by clinic for all clinics involved in the study. Please read the following chapters for more information on the results of this study:


• Chapter 3: Prevention and Screening;
  
• Chapter 4: Disease Management;
  
• Chapter 5: Healthcare Delivery;
  
• Chapter 6: Long-Term Follow-Up - presents pre- and post- PIN implementation rates for individual PIN clinics; and
  
• Chapter 7: Clinic-Based Results