Intervention to Increase Screening for Glucocorticoid Induced Diabetes (CDA-GID)

• Feasibility/Reach/Adoption [ Time Frame: 12 months ] [ Designated as safety issue: No ]
  Measurements of intervention delivery include recruitment numbers and provider Participation Rates for enrollment and retention. The definition of study feasibility consists of provider enrollment rates >= 50%. Results reported using descriptive statistics (proportions, means, standard deviations, and ranges).
• Feasibility/Reach/Adoption [ Time Frame: 12 months ] [ Designated as safety issue: No ]
  Measurements of intervention delivery include the "representativeness" of providers (differences between participants/non-participants). Results reported using descriptive statistics (proportions, means, standard deviations, and ranges).
• Feasibility/Reach/Adoption [ Time Frame: 12 months ] [ Designated as safety issue: No ]
  Measurements of intervention delivery include the rationale used by clinicians declining participation. Results reported using descriptive statistics (proportions, means, standard deviations, and ranges).
• Feasibility/Reach/Adoption [ Time Frame: 12 months ] [ Designated as safety issue: No ]
  Measurements of intervention delivery include numbers of veterans excluded from the intervention. Results reported using descriptive statistics (proportions, means, standard deviations, and ranges).

Using glucocorticoid induced diabetes (GID) we will conduct a small feasibility randomized intervention trial to improve GID management for veterans who are prescribed chronic glucocorticoids. Approximately 20 primary care providers (caring for approximately 100 veterans on chronic steroids) will receive decisional support (automatically-derived orders for hemoglobin A1C, to be co-signed by providers). Approximately 20 providers (caring for 100 veterans) who did not receive decisional support will serve as the control population. We will measure the time from randomization until the provider signs an order for hemoglobin A1C for both groups. The trial will be conducted at the VA's Eastern Colorado Health Care System (ECHCS) and will last approximately 6 months. After the completion of the trial for each provider, we will conduct a brief interview and survey of the participating providers to assess the acceptability of decisional support interventions to manage GID [protocol, page 1-2]. No patient data (PHI) will be collected by the providers.

Aim 1: Conduct a single site randomized controlled feasibility trial of a decisional support intervention to reduce ADRLLS by improving providers' adherence to GID prevention measures. Approximately 20 primary care providers (caring for approximately 100 veterans on chronic steroids who have not been screened for diabetes in the last year) will be randomly assigned to receive decisional support for GID prevention measures (automatically-derived orders for hemoglobin A1C, to be co-signed by providers). These providers will be compared with 20 providers (caring for 100 veterans not screened for diabetes in the last year) who are randomly assigned to not receive decisional support for management of GID (i.e. the control group). The trial will be conducted at the VA's Eastern Colorado Health Care System (ECHCS) and will last approximately 6 months. Study administration will be coordinated through the Colorado Research Award Enhancement Program (Colorado REAP).

Aim 1a) Procedural Endpoints: As a feasibility trial, this study includes multiple procedural endpoints, such as provider participation rates, to assess the viability of the intervention, rather than a single primary efficacy outcome measure (for details, pages 6-7).

Aim 1b) Preliminary Estimates of Efficacy: We will determine the Delay Interval—the number of days from when a provider is randomized until the provider orders the GID prevention measure. Use of this continuous measure will maximize the power of this feasibility study, though we will also determine proportions of patients for whom these measures were ordered at six months, so that we may estimate the sample size for a subsequent multi-center randomized control trial.

Aim 1c) Post-trial de-briefing: Using structured interviews, we will examine the opinions of providers after they have participated in the GID feasibility trial, in order to gather additional qualitative data regarding the ADRLL framework and refine the intervention for a subsequent randomized controlled trial. Providers will also complete a brief survey assessing their preference for the intervention.

• Behavioral: Decision support
  Clinical pharmacists mediated computerized decision support
• Behavioral: Usual Care
  clinicians typical apporach for GID monitering

• Experimental: Decision Support Intervention
  Clinical pharmacists mediated computerized decision support
  Intervention: Behavioral: Decision support
• Active Comparator: Usual Care
  Clinicians' typical approach for GID monitoring
  Intervention: Behavioral: Usual Care

Inclusion criteria:

• Must be 18-90 years old
• Must be served by the ECHCS VA sytem
• Must have chronic glucocorticoid exposure (greater than or equal to 90 days of oral glucocorticoids)
• Eligible providers will consist of those primary care practitioners within the ECHCS with at least one patient meeting the above criteria. For each veteran, the primary care provider (PCP) will be defined as the patient's current general internal medicine or family practice practitioner. In the rare instance in which the patient has no PCP within the VA system, the specialist with the greatest number of patient encounters during the past 12 months will be eligible for inclusion. For every patient, only one provider will be randomized (to avoid multiple exposure to the intervention for some patients). The consent procedure is described below.

Exclusion Criteria:

• Veterans who have had a hgbA1C test within the previous 12 months will be excluded.
• Providers without eligible patients (described above), will be excluded.
• Providers declining to give consent will be excluded.