Implementation of a peer review process to improve documentation consistency of care process indicators in the EMR in a primary care setting.

BACKGROUND: Kaiser Permanente Colorado is a group model nonprofit HMO that provides health care services to more than 500,000 members. The Primary Care Clinical Pharmacy Services (PCCPS) department consists of 33 clinical pharmacy specialists (CPS), who are located in 19 primary care clinics. OBJECTIVES: To develop and implement a peer review process to (a) improve the consistency of documentation of process indicators in the electronic medical record (EMR), (b) ensure compliance with existing standards, and (c) share best practices among PCCPS with varying geographical locations and practice styles. METHODS: A committee was formed to undertake the peer review process. An audit tool consisting of yes/no questions was created to assess chart documentation by PCCPS and to provide feedback for improvement. Four sections were included in the evaluation tool: (a) content, (b) collaborative drug therapy management, (c) nonformulary reviews, and (d) pharmacy system documentation. Peer reviews occurred quarterly, and all CPS participated. Copies of reviews were distributed to PCCPS clinicians and their supervisors. Questions and inconsistencies regarding the process were identified by the peer review committee to provide feedback to the group to optimize reviews. After completion of each quarter's reviews, error rates were calculated by dividing the total number of "no" answers by the total number of PCCPS notes reviewed that quarter. A 2-tailed Fisher's exact test was used to compare the error rate at the last quarter of each year (2007 to 2010) with baseline (2007 Q1). RESULTS: A total of 1,856 reviews were conducted between 2007 Q1 and 2010 Q2. Significant improvements in documentation were demonstrated over the first 12 months and sustained for the next 2.5 years. From 2007 Q1 to 2010 Q2, the rate of noncompliant elements decreased from 14.1% to 2.5% (P=0.001) in the content section and decreased from 31.3% to 8.3% (P less than 0.001) across all sections. CONCLUSIONS: Over 3 years of follow-up, the peer review process was successful in improving the consistency of documentation by PCCPS and compliance with existing standards. The process was well received by participants. The peer review document is easily adaptable and can be updated to address changes in drug therapy management protocols and nonformulary medication reviews as needed. This process also allows for sharing of best practices among high-functioning PCCPS practitioners who otherwise could remain isolated.

The effect of a multidisciplinary program on HAART adherence.

Although emerging evidence suggests differing interventions may improve antiretroviral adherence, there has not been a formal evaluation to identify the impact of a clinic-based multidisciplinary program designed to provide education and identify and correct potential adherence barriers prior to the initiation of highly active antiretroviral therapy (HAART). A retrospective cohort study utilizing a historical control group was conducted to compare duration on antiretrovirals, clinical indicators, and adherence rates, as captured by pharmacy refill records. Two hundred sixty-one patients met criteria for inclusion (109 subjects, 152 controls). Time to stopping antiretrovirals, as evidenced by Kaplan-Meier plot, was significantly higher in Protocol group than Controls (log-rank p = 0.023): the median duration on HAART for the intervention group was greater than 360 days but only 210 days for the control group. Thus, more subjects in the protocol group continued on therapy for the full year: 60 (55%) versus 65 (43%) for the control group. The mean reduction in log10 viral loads between HAART initiation and 12 months was greatest for the intervention group with viral load at HAART initiation 100,000 copies per milliliter or more, -3.57 versus -1.78 for controls with viral load less than 100,000 copies per milliliter (p < 0.001). For the intervention group, the mean number of adherence barriers identified per person was 4% and 72% were found to have three or more barriers. Patients at high risk for poor adherence benefit from multidisciplinary education and proactive identification of adherence barriers by exhibiting prolonged duration on therapy and greater reduction in log10 viral loads.

Clinical and economic outcomes of conversion of simvastatin to lovastatin in a group-model health maintenance organization.

OBJECTIVE: To (a) determine if converting patients on simvastatin to lovastatin affects whether they meet their low-density lipoprotein cholesterol (LDL-C) goals as defined by the National Cholesterol Education Program Adult Treatment Panel III (ATP III) clinical practice guidelines and (b) assess the change in health care expenditures associated with such a conversion. METHODS: Primary and secondary prevention patients receiving simvastatin 10 mg to 40 mg daily between September 1, 2001, and February 28, 2002, were offered lovastatin at a therapeutically equivalent dose. Fasting lipid profiles and alanine aminotransferase (ALT) levels were measured and recorded at least 6 weeks after starting lovastatin. A clinical pharmacy staff member, in collaboration with the subject's primary care provider, subsequently adjusted lipid-lowering therapy as needed to attain target LDL-C goals, as determined by the ATP III clinical practice guidelines. RESULTS: Of 5,286 patients converted to lovastatin and for whom follow-up laboratory tests were drawn, 5,046 (95.5%) were converted successfully, and 240 (4.5%) had to be converted back to simvastatin due to intolerance (N=164, 3.1%) or failure to achieve LDL-C goal (N=76, 1.4%). The proportion of patients with LDL-C at or less than their target goal increased from 75.9% before the intervention to 79.1% after conversion to lovastatin (P <0.001). ALT levels did not change significantly. The mean ALT value, a proxy measure of safety before and after conversion for all patients, was 26.9 IU/L and 26.4 IU/L, respectively (P=0.134). For the 2,235 patients converted from lovastatin 80 mg to simvastatin 40 mg, the mean pre-ALT and post-ALT values were 26.9 IU/L and 26.5 IU/L (P=0.498). The annualized cost savings due to the conversions, expressed across the entire membership of this health maintenance organization (HMO), was 4.14 US dollars per member per year (PMPY), with no change in ALT levels. Patient savings in reduced copayments in the conversion from brand simvastatin to generic lovastatin were an average of 145.29 US dollars (62%) per patient (95% confidence interval, 143-149 US dollars, P <0.001). CONCLUSION: A clinical pharmacy-directed program designed to convert patients from simvastatin to lovastatin resulted in substantial expenditure reductions for this HMO and 62% copayment savings for members, without compromise in clinical outcomes as measured by lipid control (effectiveness) and ALT levels (safety). The proportion of patients at or less than their LDL-C goal increased coincident with the conversion from simvastatin to lovastatin.

Low myopathy rates associated with statins as monotherapy or combination therapy with interacting drugs in a group model health maintenance organization.

STUDY OBJECTIVE: Because the risk for myopathy increases when 3-hydroxy-3-methylglutaryl coenzyme A inhibitors (statins) are used with other agents known to inhibit cytochrome P450 3A4 in patients with dyslipidemia, we sought to quantify this risk in a diverse, real-world sample of patients receiving statin therapy. DESIGN: Retrospective chart review. SETTING: Kaiser Permanente Colorado (KPCO), a group model health maintenance organization with approximately 360,000 members. PATIENTS: Four hundred sixty-eight patients who were identified as having a diagnosis of myopathy over a 4-year period using KPCO computerized data systems. MEASUREMENTS AND MAIN RESULTS: Medical records were reviewed to confirm myopathy cases associated with statin therapy. Of the 468 patients, 61 had received statin therapy before their diagnosis, and 41 (67%) of these patients had confirmed myopathy (documented creatine kinase level>or=1000 IU/L). The prevalence of myopathy was 0.12% with statin monotherapy and 0.22% with statins in combination with interacting drugs. Only 17 of the 41 (41%) patients had confirmed myopathy with no other plausible clinical explanation, such as a muscle injury. Increased risk of myopathy associated with statin therapy in combination with interacting drugs approached statistical significance (p=0.052) but was of minimal clinical significance. CONCLUSION: The prevalence of confirmed myopathy in patients receiving statin therapy is low (<1%). Combining statin therapy with interacting drugs (e.g., fibrates) was not associated with a clinically important increase in the prevalence of myopathy. The risk of developing myopathy during statin therapy is outweighed by the benefits derived from the therapeutic effects of the therapy.