Early steps in the development of a claims-based targeted healthcare safety monitoring system and application to three empirical examples.

BACKGROUND: Several efforts are under way to develop and test methods for prospective drug safety monitoring using large, electronic claims databases. Prospective monitoring systems must incorporate signalling algorithms and techniques to mitigate confounding in order to minimize false positive and false negative signals due to chance and bias. OBJECTIVE: The aim of the study was to describe a prototypical targeted active safety monitoring system and apply the framework to three empirical examples. METHODS: We performed sequential, targeted safety monitoring in three known drug/adverse event (AE) pairs: (i) paroxetine/upper gastrointestinal (UGI) bleed; (ii) lisinopril/angioedema; (iii) ciprofloxacin/Achilles tendon rupture (ATR). Data on new users of the drugs of interest were extracted from the HealthCore Integrated Research Database. New users were matched by propensity score to new users of comparator drugs in each example. Analyses were conducted sequentially to emulate prospective monitoring. Two signalling rules--a maximum sequential probability ratio test and an effect estimate-based approach--were applied to sequential, matched cohorts to identify signals within the system. RESULTS: Signals were identified for all three examples: paroxetine/UGI bleed in the seventh monitoring cycle, within 2 calendar years of sequential data; lisinopril/angioedema in the second cycle, within the first monitoring year; ciprofloxacin/ATR in the tenth cycle, within the fifth year. CONCLUSION: In this proof of concept, our targeted, active monitoring system provides an alternative to systems currently in the literature. Our system employs a sequential, propensity score-matched framework and signalling rules for prospective drug safety monitoring and identified signals for all three adverse drug reactions evaluated.

Validation of claims-based diagnostic codes for idiopathic thrombotic thrombocytopenic purpura in a commercially-insured population.

It was the purpose of the present study to validate administrative claims codes for idiopathic thrombotic thrombocytopenic purpura (TTP) in a commercially-insured US population. Patients with at least one medical claim with ICD-9 code 446.6X between 1/1/2001 and 5/31/2008 were identified in the HealthCore Integrated Research Database (HIRD). A chart abstraction form was developed to enable case determination for patients identified by the claims code. Two clinical experts, not involved in the design of the study, reviewed the abstracted medical record data and determined whether definite evidence supporting the diagnosis of TTP was present. The positive predictive value (PPV) of the claims coding algorithm for cases assessed by both reviewers was computed. The claims algorithm was further refined and the PPV of the refined algorithm was computed. One hundred eighty-nine abstracted charts were reviewed by two clinical experts; 86 were assessed to have definite evidence supporting the diagnosis of TTP (PPV 45.5% [86/189; 95% confidence interval (CI), 38.3-52.9%]). Refinement of the claims algorithm first included the use of plasma exchange treatment, resulting in 103 potential cases, of which 67 were assessed to have definite evidence supporting the diagnosis of TTP (PPV 65.0%; 95% CI, 55.0-74.2%). Further refinement of the claims algorithm ruled out alternative diagnoses that may mimic TTP; 34 were assessed to have definite evidence supporting the diagnosis of TTP (PPV 72.3% [34/47; 95% CI, 57.4-84.4%]).Our findings demonstrate the difficulty of confirming the diagnosis of rare disorders that lack definite diagnostic criteria, and indicate that more complex claims coding algorithms are necessary for identifying these disorders.

Clinical outcomes in sarcoidosis after cessation of infliximab treatment.

BACKGROUND AND OBJECTIVE: Infliximab appears to be efficacious for the treatment of recalcitrant forms of sarcoidosis. However, there are minimal data concerning the course of sarcoidosis once infliximab is discontinued. METHODS: Clinical outcomes in patients who had received infliximab and had discontinued it for at least 2 months were analysed retrospectively. The severity of involvement of the index organ from the time of discontinuation of infliximab was compared with that at the end of the follow-up period. RESULTS: Fourteen patients with sarcoidosis who had been treated with infliximab and had discontinued this therapy were identified. Before discontinuation of infliximab, 9 of the 14 patients (64%) responded to infliximab treatment and only one (7%) deteriorated. Patients who discontinued infliximab were followed for a mean of 12 months. At the end of the follow-up period, 12 of the 14 patients (86%) had deteriorated as compared with their status at the time of discontinuation of infliximab and two (14%) had remained stable. Kaplan-Meier analysis of time to clinical deterioration showed that half the patients deteriorated within 3 months of discontinuing infliximab. Patients who had discontinued infliximab appeared to be more likely to have their dose of prednisone increased. CONCLUSION: Patients with recalcitrant sarcoidosis who receive infliximab appear likely to deteriorate after discontinuation of this medication.

The utility of the chest radiograph in diagnosing exacerbations of pulmonary sarcoidosis.

BACKGROUND AND OBJECTIVE: To quantify the CXR using a profusion of small lung opacities score in patients with pulmonary exacerbation of sarcoidosis. In particular, the study sought to determine whether the CXR changes were a reliable indicator of disease exacerbation. METHODS: The study recruited patients with an exacerbation of pulmonary sarcoidosis, who were attending a university medical centre sarcoidosis clinic. Pulmonary exacerbation was defined as worsening pulmonary symptoms, thought to be related to sarcoidosis, that responded to an increase in corticosteroid dose. A subset of patients were identified who met the criteria of spirometric decline of >or=10% in FVC or FEV(1) from the previous (baseline) visit. Patients required a baseline CXR and spirometry and had CXR and spirometry performed at the initial assessment. Two International Labour Organisation B readers interpreted the CXR in a blinded manner. RESULTS: All study patients (n = 36) were African American; there were 24 patients in the 'spirometric decline' subgroup. Agreement between the B readers was moderate (kappa = 0.54, P < 0.01). The mean profusion score increased, or worsened by 1.38 points (SD = 3.60, P = 0.008) for the exacerbation group while the 'spirometric decline' group worsened by an average of 1.80 points (SD = 3.81, P = 0.031). There was too much variation for a cut-off to be identified that would reliably diagnose exacerbations. There was no significant correlation between FVC or FEV(1) and profusion score. CONCLUSIONS: The CXR International Labour Organisation profusion score is not a reliable test to detect pulmonary exacerbations of sarcoidosis. However, a CXR may be useful to exclude other diagnostic possibilities.

Glutathione S-transferase M1 polymorphism and the risk of lung cancer.

Glutathione S-transferases (GSTs) are a family of enzymes that detoxify hydrophobic electrophiles, including polycyclic aromatic hydrocarbon carcinogens that have been implicated in the pathogenesis of lung cancer. The GSTM1 gene within the mu class of human GSTs has been shown to be polymorphic, with individuals who are homozygous for a null allele having the GSTM1-null genotype. Individuals with the GSTM1-null genotype are deficient in both the GSTM1 and GSTM3 isoenzymes in the lung. A number of epidemiological studies have been conducted to assess the association between the GSTM1-null genotype and the risk of lung cancer. Although there have been conflicting reports regarding this relationship, the current weight of evidence indicates that the GSTM1-null genotype is probably associated with a modest increase in the risk of lung cancer. This risk appears to be greater in African-American and Asian populations than in Caucasians. Recent investigations have shown that the GSTM1-null genotype combined with CYP1A1, NAT2, or GSTP1 polymorphisms confers a greater risk of lung cancer than the GSTM1-null genotype alone. Future investigations should focus on assessing the risk of lung cancer related to multiple combinations of genetic polymorphisms that may identify individuals who are at high risk for developing lung cancer with a greater degree of certainty than is currently possible. This could lead to new clinical strategies for counseling, risk reduction and the detection of lung at an early and potentially curable stage.