Table 4 Examples of QUERI organizational research findings and their application in QUERI implementation research

Mental Health (MH) QUERI

Depression

• Guidelines adapted for local use taking organizational resources and priorities into account

• Assessed national sample of PC clinics to understand variations in structure and processes of care (e.g., PC-based vs. referral focus)

• Used national organizational survey to measure factors associated with PC-MH joint management of depression:

• practice size (small-to-medium size)

• more generalist MDs (vs. MD extenders)

• greater specialty access (vs. pre-authorization for specialty use)

• higher PC practice autonomy and provider incentives

• Guidelines updated based on lessons learned from new randomized trials (Steps #4–6 full circle to revise Step #1)

• Used knowledge of organizational factors (Step #3) to select 1^st generation sites for implementing collaborative care (e.g., small-to-medium size sites with evidence of joint PC-MH management)

• Measured site-specific structure using interviews of PC and MH leaders

• Used site variations to target additional intervention resources to sites needing more provider education to ensure formulary access to antidepressants

• Adapted intervention to accommodate staffing constraints (e.g., use of telephone vs. on-site care manager)

• Identified organizational factors associated with adoption/penetration of collaborative care (e.g., sites with greater autonomy tend to push intervention to more providers faster but have greater difficulty sustaining it than sites that take more time to adapt the intervention among smaller provider groups).

• Applied organizational factors to further adapt implementation for rollout to 2^nd generation sites

Substance Use Disorders QUERI

Smoking cessation

• Used national organizational survey to measure factors associated with higher tobacco counselling rates:

• small-to-medium non-academic VAs

• sites with greater staff commitment to QI

• sites with integrated nurse practitioners and behavioural health professionals in PC practice

• sites with separate PC budgets

• sites with inpatient-outpatient continuity

• Used site surveys and administrative data to ascertain organizational resources before introducing evidence-based options (e.g., PC-based changes in care vs. specialty referral-based changes)

• Used organizational factors to pair PC practices on size and academic affiliation in group randomized trial

• Measured site-specific structure during and after implementation using key informant organizational surveys

• Adapted intervention to accommodate local structural variations (e.g., added pharmacotherapy training)

• Redesigned intervention to address factors that hindered adoption (e.g., telephone counselling)

Alcohol use disorders

• Used national organizational survey to evaluate factors associated with PC management of alcohol use:

• sufficiency of PC clinical support arrangements

• physician involvement in QI

• statistician for decision support

• PCP responsibility for chronic care

• availability of seminars on cost-effective care

• Combined organizational surveys of VA primary care practices and substance use programs to evaluate availability of alcohol treatment programs

• Further organizational research planned before design and implementation of QI interventions

Colorectal Cancer QUERI

Colorectal cancer (CRC) screening

• Measured system capacity for colonoscopy using key informant organizational survey:

• availability of/access to GI specialists

• key coordination mechanisms between PC-GI needed

• Used national organizational survey to evaluate factors associated with higher CRC screening rates:

• PC practice autonomy

• sufficiency of clinical practice support arrangements in PC practice

• smaller PC practices

• Implementation of new organizational supports for obtaining colonoscopies for patients with +FOBT

• Evaluated interaction between organizational and patient-level factors (e.g., racial-ethnic/gender differences)

• Measured CRC-specific organizational factors (e.g., GI staffing, use of PC-GI service agreements, use of community providers) to inform intervention design

• Integrated GI staffing and other organizational variables into system-level VA cost-effectiveness model

HIV/Hepatitis QUERI

HIV disease

• Categorized VA facilities based on:

• HIV caseload

• Use of HIV guidelines

• Methods of promoting adherence (e.g., chart audits, feedback)

• Used national HIV organizational survey to measure HIV care variations:

• Most urban VAs have special HIV clinics staffed with experienced HIV providers; rural VAs tend to manage HIV in PC, use outside experts

• Most VAs have 1+ HIV case manager

• Used national organizational survey to measure organizational readiness for change, local barriers and preferences for different types of QI implementation

• Used organizational care arrangements from national survey to select sites for trial (i.e., minimum eligibility criteria) (e.g., adopted HIV QI guidelines, reported provider readiness for change)

• Evaluated organizational factors associated with adoption of HIV guidelines (e.g., urban, complex, larger HIV caseloads, use HIV case managers, fewer barriers to antiretroviral therapy and opportunistic infection prophylaxis guidelines) and HIV-related QI (e.g., larger, more complex facilities)

• Used administrative data to classify VA facilities by level of organizational attributes of HIV care and analyzed links to better control of HIV infection

Diabetes QUERI

Diabetes mellitus

• Used organizational surveys to benchmark VA practices with those outside the system

• Appraised performance variations at the patient, provider and facility levels

• Used organizational surveys to identify factors associated with glycemic control:

• Greater PC authority over establishing clinical policies

• Greater staffing authority

• Greater use of computerized diabetes reminders

• Special teams or protocols to respond to clinical issues

• Weekly multidisciplinary clinical team meetings

• Used PC provider survey to study influences of organization of care and provider training on treatment of pain among diabetics (e.g., inadequate training in chronic pain management, treatment of pain conditions perceived as beyond provider's scope of experience)

• Evaluating clinician, organizational and patient factors contributing to failure to change therapy when blood pressure among diabetics is elevated