Improving Diabetes Care:Effectiveness of Physician Profiling and Care Coordination by a Diabetes Resource Nurse

• Change Score for "HbA1c <9 Percent" [ Time Frame: This measure compared baseline values (01/01/2000-12/31/2000) to follow-up values (01/01/2001-12/31/2001) ] [ Designated as safety issue: No ]
  Each patient was assigned a "change score" of -1, 0, or 1. A positive value indicated a patient non-adherent to the guideline recommendation for HbA1c <9 percent at baseline had achieved such a level at follow up. Patient-level change scores were then summed and averaged over each study arm.
• Change Score for "LDL <100 mg/dL" [ Time Frame: change from baseline (01/01/2000-12/31/2000) to follow-up (01/01/2001-12/31/2001) ] [ Designated as safety issue: No ]
  Each patient was assigned a "change score" of -1, 0, or 1. A positive value indicated that a patient non-adherent to the guideline recommendation of LDL <100 mg/dL at baseline had achieved adherence at follow-up. Patient-level change scores were then summed and averaged over each study arm.
• Change Score for "Blood Pressure (b.p.) <130/80 mmHg" [ Time Frame: change from baseline (01/01/2000-12/31/2000) to follow-up (01/01/200112/31/2001) ] [ Designated as safety issue: No ]
  Each patient was assigned a change score of -1, 0, or 1. A positive value indicated that a patient non-adherent to the guideline recommendation of blood pressure <130/80 mmHg at baseline had achieved adherence at follow-up. Patient-level change scores were then summed and averaged over each study arm.

• Change Score for "HbA1c Level" [ Time Frame: change from baseline (1/1/2000-12/31/2000) to follow-up (1/1/2001-12/31/2001) ] [ Designated as safety issue: No ]
  Change score was calculated by subtracting the follow-up HbA1c value from the baseline value for each patient. Patients missing either a baseline or follow-up value were excluded. Patient-level change scores were then summed and averaged over each study arm.
• Change Score for "Diastolic Blood Pressure" [ Time Frame: change from baseline (1/1/2000-12/31/2000) to follow-up (1/1/2001-12/31/2001) ] [ Designated as safety issue: No ]
  Change score was calculated by subtracting the follow-up value from the baseline value for each patient. Patients missing either a baseline or follow-up value were excluded. Patient-level change scores were then summed and averaged over each study arm.
• Change Score for "LDL Level" [ Time Frame: change from baseline (1/1/2000-12/31/2000) to follow-up (1/1/2001-12/31/2001) ] [ Designated as safety issue: No ]
  Change score was calculated by subtracting the follow-up LDL value from the baseline value for each patient. Patients missing either a baseline or follow-up value were excluded. Patient-level change scores were then summed and averaged over each study arm.
• Change Score for "Systolic Blood Pressure" [ Time Frame: change from baseline (1/1/2000-12/31/2000) to follow-up (1/1/2001-12/31/2001) ] [ Designated as safety issue: No ]
  Change score was calculated by subtracting the follow-up value from the baseline value for each patient. Patients missing either a baseline or follow-up value were excluded. Patient-level change scores were then summed and averaged over each study arm.
• Change Score for "Annual HbA1c Assessment" (as Determined From Medical Record Review) [ Time Frame: change from baseline (1/1/2000-12/31/2000) to follow-up (1/1/2001-12/31/2001) ] [ Designated as safety issue: No ]
  Each patient was assigned a change score of -1, 0, or 1. A positive value indicated a patient who was non-compliant with the guideline recommendation for an annual HbA1c assessment in the baseline period was compliant in the follow-up period. Patients missing either a baseline or follow-up value were excluded. Patient-level change scores were then summed and averaged over each study arm.
• Change Score for "Annual Lipid Assessment" (as Determined From Medical Record Review) [ Time Frame: change from baseline (1/1/2000-12/31/2000) to follow-up (1/1/2001-12/31/2001) ] [ Designated as safety issue: No ]
  Each patient was assigned a change score of -1, 0, or 1. A positive value indicated a patient who was non-compliant with the guideline recommendation for an annual lipid assessment in the baseline period was compliant in the follow-up period. Patients missing either a baseline or follow-up value were excluded. Patient-level change scores were then summed and averaged over each study arm.
• Change Score for "Annual Blood Pressure Assessment" (as Determined From Medical Record Review) [ Time Frame: change from baseline (1/1/2000-12/31/2000) to follow-up (1/1/2001-12/31/2001) ] [ Designated as safety issue: No ]
  Each patient was assigned a change score of -1, 0, or 1. A positive value indicated a patient who was non-compliant with the guideline recommendation for an annual blood pressure assessment in the baseline period was compliant in the follow-up period. Patients missing either a baseline or follow-up value were excluded. Patient-level change scores were then summed and averaged over each study arm.
• Change Score for "Annual Foot Exam" (as Determined by Medical Record Review) [ Time Frame: change from baseline (1/1/2000-12/31/2000) to follow-up (1/1/2001-12/31/2001) ] [ Designated as safety issue: No ]
  Each patient was assigned a change score of -1, 0, or 1. A positive value indicated a patient who was non-compliant with the guideline recommendation for an annual foot exam in the baseline period was compliant in the follow-up period. Patients missing either a baseline or follow-up value were excluded. Patient-level change scores were then summed and averaged over each study arm.
• Change Score for "Annual Eye Exam" (as Determined From Medical Record Review) [ Time Frame: change from baseline (1/1/2000-12/31/2000) to follow-up (1/1/2001-12/31/2001) ] [ Designated as safety issue: No ]
  Each patient was assigned a change score of -1, 0, or 1. A positive value indicated a patient who was non-compliant with the guideline recommendation for an annual eye exam in the baseline period was compliant in the follow-up period. Patients missing either a baseline or follow-up value were excluded. Patient-level change scores were then summed and averaged over each study arm.
• Change Score for "Annual Renal Function Assessment" (as Determined From Medical Record Review) [ Time Frame: change from baseline (1/1/2000-12/31/2000) to follow-up (1/1/2001-12/31/2001) ] [ Designated as safety issue: No ]
  Each patient was assigned a change score of -1, 0, or 1. A positive value indicated a patient who was non-compliant with the guideline recommendation for an annual renal function assessment in the baseline period was compliant in the follow-up period. Patients missing either a baseline or follow-up value were excluded. Patient-level change scores were then summed and averaged over each study arm.
• Change Score for "Annual HbA1c Assessment" (as Determined From Medicare Claims Data) [ Time Frame: change from baseline (1/1/2000-12/31/2000) to follow-up (1/1/2001-12/31/2001) ] [ Designated as safety issue: No ]
  Each patient was assigned a change score of -1, 0, or 1. A positive value indicated a patient who was non-compliant with the guideline recommendation for an annual HbA1c assessment in the baseline period was compliant in the follow-up period. Patients missing either a baseline or follow-up value were excluded. Patient-level change scores were then summed and averaged over each study arm.
• Change Score for "Annual Eye Exam" (as Determined From Medicare Claims Data) [ Time Frame: change from baseline (1/1/2000-12/31/2000) to follow-up (1/1/2001-12/31/2001) ] [ Designated as safety issue: No ]
  Each patient was assigned a change score of -1, 0, or 1. A positive value indicated a patient who was non-compliant with the guideline recommendation for an annual eye exam in the baseline period was compliant in the follow-up period. Patients missing either a baseline or follow-up value were excluded. Patient-level change scores were then summed and averaged over each study arm.
• Change Score for "Annual Lipid Assessment" (as Determined From Medicare Claims Data) [ Time Frame: change from baseline (1/1/2000-12/31/2000) to follow-up (1/1/2001-12/31/2001) ] [ Designated as safety issue: No ]
  Each patient was assigned a change score of -1, 0, or 1. A positive value indicated a patient who was non-compliant with the guideline recommendation for an annual lipid assessment in the baseline period was compliant in the follow-up period. Patients missing either a baseline or follow-up value were excluded. Patient-level change scores were then summed and averaged over each study arm.
• Change Score for "Semiannual HbA1c Assessment" (as Determined From Medicare Claims Data) [ Time Frame: change from baseline (1/1/2000-12/31/2000) to follow-up (1/1/2001-12/31/2001) ] [ Designated as safety issue: No ]
  Each patient was assigned a change score of -1, 0, or 1. A positive value indicated a patient who was non-compliant with the guideline recommendation for a semi-annual HbA1c assessment in the baseline period was compliant in the follow-up period. Patients missing either a baseline or follow-up value were excluded. Patient-level change scores were then summed and averaged over each study arm.

The purpose of this study is to test the effectiveness of physician profiling and care coordination by a diabetes resource nurse in improving the quality of diabetes care.

HealthTexas Provider Network primary care practices with at least 10 Medicare diabetes patients over the age of 65 were randomized to one of 3 intervention arms: physician feedback of process measures using Medicare claims data ("Claims"); feedback of Medicare claims data plus clinical measures from medical record abstraction ("Claims+MR"); or both types of feedback plus a practice-based DRN ("DRN"). For the 12 months prior to the intervention and 12 months post-intervention, performance data on diabetes related processes of care (annual HbA1c testing, annual LDL cholesterol screening, annual hypertension screening, annual eye, foot, and renal assessment) and patient outcomes (HbA1c level, LDL cholesterol level, blood pressure) were collected from medical record abstraction and Medicare claims data. Pre-post change scores will be compared between intervention arms to examine effectiveness of physician profiling and care coordination by a diabetes resource nurse.

• Other: Medicare Claims Feedback
  Physician practices received periodic feedback on their performance on selected diabetes quality of care measures as reflected by the Medicare claims data for their patients.
• Other: Medical Record Review
  Physician practices received period feedback on their performance on selected diabetes quality of care measures, as reflected by data collected from their patients' medical records. These data were compiled by trained nurse abstractors using a standardized data collection tool developed for this study.
• Other: Diabetes Resource Nurse
  Diabetes Resource Nurses (DRNs) were registered nurses with 3-5 years of experience as certified diabetes educations who performed initial patient assessments, developed plans of care, administered screening tools, and monitored clinical outcomes. Physicians at the practices randomised to this intervention had could access the DRN's services for their diabetes patients, but neither physicians nor patients had to take advantage of this resource.

• Active Comparator: Medicare Claims Feedback
  Practices randomised to the Medicare Claims Feedback arm received period feedback on their performance on selected diabetes quality of care measures as reflected in the claims data for their diabetes patients.
  Intervention: Other: Medicare Claims Feedback
• Experimental: Medicare Claims+Medical Record Feedback
  Practices randomised to the Medicare Claims + Medical Record Review Feedback arm received periodic feedback on their performance on selected diabetes quality of care measures as reflected in both the Medicare claims for the diabetes patients AND review/audit of their diabetes patients' medical records.
  Interventions:

  • Other: Medicare Claims Feedback
  • Other: Medical Record Review
• Experimental: Medicare Claims+Medical Chart Review+DRN
  In addition to the performance data from both Medicare Claims data and from review of patients' medical records, practices randomised to the Medicare Claims + Medical Record review + Diabetes Resource Nurse (DRN) had a diabetes resource nurse assigned to them, who was available to provide diabetes education and care-coordination type services for their diabetes patients.
  Interventions:

  • Other: Medicare Claims Feedback
  • Other: Medical Record Review
  • Other: Diabetes Resource Nurse

• Hollander P, Nicewander D, Couch C, Winter D, Herrin J, Haydar Z, Ballard DJ. Quality of Care of Medicare Patients With Diabetes in a Metropolitan Fee-for-Service Primary Care Integrated Delivery System. Am J Med Qual. 2005 Nov-Dec;20(6):344-52.
• Herrin J, Nicewander DA, Hollander PA, Couch CE, Winter FD, Haydar ZR, Warren SS, Ballard DJ. Effectiveness of diabetes resource nurse case management and physician profiling in a fee-for-service setting: a cluster randomized trial. Proc (Bayl Univ Med Cent). 2006 Apr;19(2):95-102.

Inclusion Criteria:

• age ≥ 65 years on January 1, 2000
• diagnosis of diabetes mellitus
• diabetes related visit to HTPN physician within the past year
• Resident of Texas
• Medicare insurance coverage

Exclusion Criteria:

• Patient chart not available for abstraction