External validation of the In-hospital Mortality for PulmonAry embolism using Claims daTa (IMPACT) multivariable prediction rule.

OBJECTIVE: To validate the In-hospital Mortality for PulmonAry embolism using Claims daTa (IMPACT) multivariable prediction rule using admission claims data. STUDY DESIGN: Retrospective claims database analysis. METHODS: This analysis was performed using Humana admission claims data from January 2007 to March 2014. We included adult patients admitted for their first PE during this period (International Classification of Diseases, ninth edition, Clinical Modification code of 415.1x in in the primary position or secondary position when accompanied by a primary code for a PE complication). The IMPACT rule, consisting of age plus 11 comorbidities, was used to estimate patients' probability of in-hospital mortality and classify risk. Low risk was defined as in-hospital mortality ≤ 1.5%. IMPACT was evaluated by evaluating prognostic test characteristic values and 95% confidence intervals (CIs). RESULTS: A total of 23,858 patients admitted for PE were included, and 3.3% died in-hospital. The IMPACT prediction rule classified 2371 (9.9%) as low-risk; with a sensitivity of 97.6%, 95% CI: 96.1-98.5, specificity of 10.2%, 95% CI: 9.8-10.6, negative and positive predictive values of 99.2% (95% CI: 98.7-99.5) and 3.5% (95% CI: 3.3-3.8) and c-statistic of 0.70, 95% CI: 0.0.68-0.72, for in-hospital mortality. IMPACT classified 42.7% of patients < 65 years old as low-risk; with a sensitivity, specificity and c-statistic of 85.0%, 95% CI: 77.4-90.5, 43.3%, 95% CI: 42.0-44.7 and 0.74, 95% CI: 0.69-0.78, respectively. CONCLUSION: The IMPACT prediction rule was valid when implemented in a database consisting largely of Medicare claims. Following further external validation and direct comparison to commonly used clinical prediction rules, IMPACT may become a valuable tool for payers and hospitals wishing to retrospectively assess whether their PE patients are being kept hospitalized for the optimal period of time.

Incidence of hospital readmission in patients diagnosed with DVT and PE: clinical burden of recurrent events.

BACKGROUND: Venous thromboembolism (VTE), which comprises deep-vein thrombosis (DVT) and pulmonary embolism (PE), is associated with significant morbidity and mortality and represents a considerable economic burden to the US healthcare system. Although it is well established that patients with an initial VTE are at increased risk for recurrent VTE, limited data exist on the clinical burden of a secondary DVT or PE event. The objective of this retrospective observational study was to conduct an epidemiologic evaluation, from a hospital perspective, of patients with an initial DVT or PE who experienced a recurrent event postdischarge requiring hospital readmission. METHODS: Hospital claims containing DVT or PE as a primary diagnosis for hospitalisation during the period October 2009 to April 2013 were identified by retrospective analysis using the MarketScan database. The time to hospital readmission for DVT or PE was assessed using the MarketScan Treatment Pathways tool. RESULTS: Of 214,901 patient admissions identified with a diagnosis of DVT or PE at hospital admission, approximately 4% were subsequently readmitted to the hospital with a diagnosis of PE (8217) or DVT (9138). Of all readmitted patients with a diagnosis of DVT on initial admission, 66% were rehospitalised with a diagnosis of DVT, and 34% were rehospitalised with a diagnosis of PE. Of all readmitted patients with a diagnosis of PE on initial admission, 63% were rehospitalised with a diagnosis of PE and 37% with a diagnosis of DVT. Of all hospital readmissions with a diagnosis of PE or DVT, 62% and 58% occurred within the first 30 days following an initial PE or DVT event, respectively. CONCLUSIONS: The burden of DVT or PE is large, not only because of the initial hospitalisation event but also because of the high number of hospital readmissions, more than half of which occur within 30 days.

Prasugrel vs. clopidogrel in acute coronary syndrome patients treated with prasugrel.

WHAT IS KNOWN AND OBJECTIVE: Current guidelines recommend a combination of clopidogrel and aspirin for management of patients who have experienced an acute coronary syndrome (ACS). Additional antiplatelet agents have been recently approved. Few comparative effectiveness studies are available for these new agents. Accordingly, we evaluated effect on time to hospital admission and resource utilization (number of hospitalizations, ER visits and outpatient visits) of prasugrel vs. clopidogrel in prasugrel-treated patients as assessed in a matched cohort. METHODS: Based on the Truven Health Analytics MarketScan database from 01 January 2009 through 31 July 2012, a retrospective prasugrel-clopidogrel matched cohort was created. Inferences for average treatment effect over 1 and 12 months on time to hospitalization and resource utilization were performed by (i) frequentist Kaplan-Meier estimation with a Cox proportional hazard model and Lin's cost history method for censored resource utilization outcomes and (ii) Bayesian discrete-time hazard and negative binomial models. RESULTS AND DISCUSSION: The 10,963 matched pairs were well balanced on baseline characteristics. Frequentist analyses of time to hospital admission over 365 days and mean all-cause resource utilization over 30 and 365 days showed no statistical differences between prasugrel and clopidogrel (P-values > 0·05). Based on Bayesian analysis of time to admission over 12 months, there was positive evidence of equivalence (0·987 probability of equivalence at a 10% equivalence margin and a Bayes factor of 0·611). Although the frequentist analyses for number of all-cause hospitalizations showed a lack of a significant difference at Months 1 and 12, the Bayesian data analysis showed positive evidence of superiority of clopidogrel at Month 1 (Bayes factor: 5·369); however, at Month 12, there was little evidence of superiority of one treatment over the other (Bayes factor: 0·422). WHAT IS NEW AND CONCLUSION: Using frequentist and Bayesian data analyses, in prasugrel-treated patients, clopidogrel was equivalent to prasugrel for time to hospital admission over 12 months and there was positive evidence that it was superior to prasugrel for number of hospitalizations over the first month of treatment.

Adherence to oral second-generation antipsychotic medications in patients with schizophrenia and bipolar disorder: physicians' perceptions of adherence vs. pharmacy claims.

OBJECTIVE: To compare physician-reported adherence of specific patients to oral second-generation antipsychotics vs. actual adherence rates determined from the patients' pharmacy claims. METHODS: Claims data from the HealthCore Integrated Research Database identified patients with schizophrenia or bipolar disorder with ≥ 1 oral second-generation antipsychotic prescription. The prescribing physicians were identified from the pharmacy claims and asked to complete an Internet survey assessing their perception of medication adherence for 1-2 of their patients and their beliefs regarding adherence to second-generation antipsychotics in general for a 1-year period. Adherence to second-generation antipsychotics was determined for each patient by pharmacy claims for the same period. Physician survey data were merged with patient claims data via unique patient identifiers, and physician-reported adherence rates were compared with claims-based rates as measured by the medication possession ratio. RESULTS: One hundred and fifty-three physicians responded to the survey, representing 214 patients (44 with claims for schizophrenia, 162 with bipolar disorder, 8 with claims for bipolar disorder and schizophrenia). Most physicians (60%) had no formal adherence training. More than two-thirds (68%) reported emphasising the importance of adherence and reported approximately 76% of their patients were adherent (≥ 71% of the time). In the schizophrenia group, 16 of 17 (94%) patients with low-to-moderate (≤ 70%) adherence levels had high (≥ 71%) physician-estimated adherence. In the bipolar disorder group, 62 of 92 (67%) patients with low-to-moderate adherence levels had high physician-estimated adherence. CONCLUSIONS/INTERPRETATION: These analyses suggest that, even when physicians are asked about specific patients in their practice, there is discordance between physician perceptions and adherence as measured through pharmacy claims. This disparity may delay appropriate interventions, potentially contributing to relapses.

Evaluation of health services use following the initiation of risperidone long-acting therapy among schizophrenia patients in the veterans health administration.

WHAT IS KNOWN AND OBJECTIVE: The introduction of long-acting injection antipsychotic agents has been associated with better treatment persistence and better subsequent patient outcomes. However, limited empirical data are available on patient outcomes resulting from the initiation of long-acting injectable antipsychotic agents. In this study, we assessed patterns of health-care utilization following the initiation of risperidone long-acting therapy (RLAT), the first and only second generation long-acting injectable antipsychotic agent, in schizophrenia patients within the Veterans Health Administration. METHODS: Patients were identified if they initiated RLAT between 1 October 2005 and 30 September 2006, were ≥ 18 years of age at the time of initiation, and had at least four injections following the initiation. Paired t-tests and McNemar tests were used to compare patterns of health services use during 12 months pre- and post-initiation. RESULTS AND DISCUSSION: Among 924 eligible study subjects, about 94% were male with mean age of 51·1 years and as high as 60% had >3 and 29% had >5 comorbid conditions. The initiators of RLAT had an average of 17·3 (SD ± 9·7) injections within the 12 months following the initiation, with an average of 14 days between injections. Between the pre- and post-initiation periods, although the number of psychiatric-related outpatient visits increased from 24·6 to 39·1 (P < 0·001), the number of psychiatric hospitalizations decreased from 1·4 to 1·0 (P < 0·001) with an average length of stay reducing from 20 to 14 days (P < 0·001). The percentage of patients who experienced at least one or two psychiatric-related hospitalizations decreased from 68·9% to 45·7% (P < 0·001) and from 34·9% to 24·4% (P < 0·001), respectively. WHAT IS NEW AND CONCLUSION: Despite the values of RLAT in treating patients with schizophrenia, RLAT is largely underutilized in routine clinical practice. This observation highlights the importance for future research to ascertain the cost-effectiveness of initiating RLAT, especially the extent to which medication adherence influences the prescription pattern of RLAT and subsequent costs of initiating RLAT.

The incremental costs of recommended therapy versus real world therapy in type 2 diabetes patients.

BACKGROUND: The goals of diabetes management have evolved over the past decade to become the attainment of near-normal glucose and cardiovascular risk factor levels. Improved metabolic control is achieved through optimized medication regimens, but costs specifically associated with such optimization have not been examined. OBJECTIVE: To estimate the incremental medication cost of providing optimal therapy to reach recommended goals versus actual therapy in patients with type 2 diabetes. METHODS: We randomly selected the charts of 601 type 2 diabetes patients receiving care from the outpatient clinics of Massachusetts General Hospital March 1, 1996-August 31, 1997 and abstracted clinical and medication data. We applied treatment algorithms based on 2004 clinical practice guidelines for hyperglycemia, hyperlipidemia, and hypertension to patients' current medication therapy to determine how current medication regimens could be improved to attain recommended treatment goals. Four clinicians and three pharmacists independently applied the algorithms and reached consensus on recommended therapies. Mean incremental medication costs, the cost differences between current and recommended therapies, per patient (expressed in 2004 dollars) were calculated with 95% bootstrap confidence intervals (CIs). RESULTS: Mean patient age was 65 years old, mean duration of diabetes was 7.7 years, 32% had ideal glucose control, 25% had ideal systolic blood pressure, and 24% had ideal low-density lipoprotein cholesterol. Care for these diabetes patients was similar to that observed in recent national studies. If treatment algorithm recommendations were applied, the average annual medication cost/patient would increase from 1525 to 2164 dollars. Annual incremental costs/patient increased by 168 dollars (95% CI 133-206 dollars) for antihyperglycemic medications, 75 dollars (57-93 dollars) for antihypertensive medications, 392 dollars (354-434 dollars) for antihyperlipidemic medications, and 3 dollars (3-4 dollars) for aspirin prophylaxis. Yearly incremental cost of recommended laboratory testing ranged from 77-189 dollars/patient. LIMITATIONS: Although baseline data come from the clinics of a single academic institution, collected in 1997, the care of these diabetes patients was remarkably similar to care recently observed nationally. In addition, the data are dependent on the medical record and may not accurately reflect patients' actual experiences. CONCLUSION: Average yearly incremental cost of optimizing drug regimens to achieve recommended treatment goals for type 2 diabetes was approximately 600 dollars/patient. These results provide valuable input for assessing the cost-effectiveness of improving comprehensive diabetes care.