Intranasal Fluticasone Propionate Observational Cohort Safety Studies: Reviewing Evidence from Databases on Two Continents.

PURPOSE: Our objective was to highlight the importance of database selection in observational research and to determine the incidence of corticosteroid-related events in patients exposed to fluticasone propionate intranasal spray (FPNS) compared with other intranasal steroids (INS). METHODS: After a feasibility study using an electronic medical record database in the UK (1990-2002), a retrospective cohort study was conducted using a large administrative claims database in the USA from 1994 to 2002 comparing the incidence and rate ratios of steroid-related events among intermittent, sub-chronic, and chronic FPNS use and other INS use episodes. RESULTS: Most patients used INS intermittently; power was low to evaluate risk associated with chronic use. Significantly elevated adjusted rate ratios were observed in the US study comparing FPNS with other INS for hypercorticism, sinusitis, abscess, and empyema, as well as a significantly decreased rate ratio for cataracts. The US claims database provided greater granularity on covariates and markers of severity to improve control of confounding for this study and time period, but neither database was able to assess the indication for prescription and the UK study could not address the use of INS without a prescription. CONCLUSIONS: The FPNS results were consistent with the risk profile for INS and did not raise any new safety signals at the time of study conduct, which is consistent with the current safety profile. We were not able to discern the extent of potential off-label use of FPNS or other INS. Differences in the available data and healthcare systems highlight important considerations for database selection in the feasibility phase to assess the precision and limitations prior to formal risk evaluation.

Epidemiology and healthcare utilization of neurogenic bladder patients in a US claims database.

AIMS: To characterize the patient profile, medication utilization, and healthcare encounters of patients with neurogenic bladder dysfunction related to incontinence. METHODS: Medical and pharmacy claims were retrospectively analyzed from April 1, 2002 to March 31, 2007 to characterize neurogenic bladder patients. There were 46,271 patients in the Neurogenic bladder cohort, and 9,315 and 4,168 patients in Multiple Sclerosis (MS) and Spinal Cord Injury (SCI) subcohorts, respectively. Demographic data, concomitant diseases, use of overactive bladder (OAB) oral drug, and healthcare encounters were summarized using descriptive statistics. RESULTS: The mean age of neurogenic bladder patients was 62.5 (standard deviation 19.6) years. A high frequency of lower urinary tract infections (UTIs; 29%-36%), obstructive uropathies (6%-11%), and urinary retention (9%-14%), was observed. Overall, 33,100 (71.5%) patients were taking an OAB oral drug; 10,110 (30.5%) patients discontinued and did not restart. During the one-year follow-up period, 39.0% (8,034) of neurogenic bladder patients had a urology visit, 31.7% (14,679) had a neurology visit, 33.3% (15,415) were hospitalized, and 14.4% (6,646) were in a nursing home (highest rates observed in SCI subcohort). UTI diagnoses comprised over 20% of all hospitalizations one-year post-index. Annually, neurogenic bladder patients averaged 16 office and 0.5 emergency room visits. CONCLUSIONS: This is the largest observational study conducted to address the epidemiology of the neurogenic bladder population, including healthcare utilization. These data suggest that patients with neurogenic bladder may have suboptimal management, indicated by high incidences of urinary tract complications and hospitalizations.

Prevalence of pediatric dyslipidemia: comparison of a population-based claims database to national surveys.

OBJECTIVE: To determine the prevalence of pediatric dyslipidemia in a large US medical insurance claims database and to compare the resulting estimate to the prevalence from the National Health and Nutrition Examination Survey (NHANES). PATIENTS AND METHODS: Children 10-18 years old who had laboratory-defined dyslipidemia were identified from the Integrated Healthcare Information Services (IHCIS) database 2003-2006. For comparison purposes, the corresponding prevalence among the US children of same age was estimated from the NHANES 1999-2004 data. RESULTS: Among the 273 064 children with at least one laboratory lipid value in the IHCIS database, 22.9% (n = 62 451) had laboratory-defined dyslipidemia. This prevalence was the same as the NHANES estimate (23.9%, 95%CI: 21.6-26.3). Elevated triglyceride (TG) was the most common type of dyslipidemia, detected among 13.2% of the IHCIS children and 14.2% of the US children, followed by elevated total cholesterol (TC), 7.7 and 9.6%, respectively. Among IHCIS dyslipidemic children, older teenage boys had higher rates than younger boys for high-density lipoprotein cholesterol (HDL-C) abnormality, but lower rates for elevated TC and low-density lipoprotein cholesterol (LDL-C). These age-related trends were also seen among NHANES dyslipidemic children. CONCLUSIONS: Analyses of a population-based claims database revealed the same prevalence of pediatric dyslipidemia as that among the US children assessed in the NHANES data. Among dyslipidemic children in the claims database, the occurrence of specific dyslipidemias appeared to vary by age and gender, a trend that was also seen among the dyslipidemic children in the US.

Longitudinal study on pediatric dyslipidemia in population-based claims database.

PURPOSE: To examine the rate of lipid testing among children from a large US medical insurance claims database, describe the characteristics of pediatric dyslipidemia, and assess the sensitivity of the International Classification of Disease, Ninth Revision, Clinical Modification (ICD-9-CM) codes for identifying dyslipidemic children. METHODS: This retrospective cohort study used the claims data from the Integrated Healthcare Information Services (IHCIS), for the years 2003-2006. Two study cohorts consisted of children with laboratory-defined and diagnosis/treatment-defined dyslipidemia, respectively. They were compared to age- and gender-matched children without dyslipidemia, with respect to co-morbidities during the 6-month prior to and 12-month after the first dyslipidemic laboratory value or diagnosis/treatment. RESULTS: Seven per cent of the children who had laboratory values available in the database had a cholesterol test during the study period. Only 15% of laboratory-defined children (n = 23,475) had a dyslipidemia diagnosis. Cholesterol-modifying medications were rarely prescribed. Substantially more laboratory-defined children than their comparators were obese (8 times), had diabetes mellitus (10 times), or had hypertension (5 times). These co-morbidities were even higher among diagnosis/treatment-defined children. CONCLUSIONS: The rate of lipid testing among children was low. The ICD-9-CM diagnostic codes showed low sensitivity against laboratory definitions. Though only a small proportion of dyslipidemic children were diagnosed or treated with a medication, co-morbidities associated with dyslipidemia were common.

Effectiveness of rosuvastatin compared to other statins for the prevention of cardiovascular events-a cohort study in 395 039 patients from clinical practice.

PURPOSE: This study compared the effectiveness of rosuvastatin (RSV) to other statins prescribed in clinical practice in prevention of cardiovascular (CV) events. METHODS: This longitudinal inception cohort study, using Thomson Healthcare's MarketScan databases, included patients aged > or = 18 starting statin therapy during August 2003-December 2005. Patients were followed until 90 days after index statin monotherapy exposure, start of another lipid-lowering therapy, an event, end of eligibility, or end of study. The primary endpoint was a composite of CV death (in-hospital only), myocardial infarction, unstable angina, coronary revascularization, stroke, and carotid revascularization. Adjusted time-to-event analyses incorporating a propensity score covariate were used, and analyses were stratified by duration of statin exposure. RESULTS: Among 395 039 patients who met inclusion/exclusion criteria, 12% initiated RSV, and 9622 (2.4%) of the total patient population experienced an outcome event. The median duration of statin treatment and follow-up was 100 days and 180 days, respectively. No statistically significant difference in CV event rates between RSV and other statins was observed after adjustment for demographics and medical/prescription history (HR = 0.99, 95%CI = 0.93-1.06). However, with longer exposure time, there was a suggestion of increased benefit with RSV compared to other statins. CONCLUSIONS: The primary analysis showed similar incidence rates of CV-related events between the statin cohorts over a median of 180 days of follow-up.

Risk of further decline in renal function after the use of oral sodium phosphate or polyethylene glycol in patients with a preexisting glomerular filtration rate below 60 ml/min.

OBJECTIVES: The aim of this study was to estimate the risk of further creatinine increase in patients with preexisting renal disease after the use of oral sodium phosphate (OSP) versus polyethylene glycol (PEG), and to study usage patterns of OSP in relation to renal function. METHODS: A cohort study was done using clinical records and electronic patient information from the Henry Ford Health System (HFHS) in patients who had used either OSP or PEG for colonoscopy between February 1999 and April 2006. Among patients with an estimated GFR < 60 ml/min before colonoscopy, we identified cases with an unexplained creatinine increase of > or = 0.5 mg/dl within 14 days after colonoscopy. RESULTS: We identified 7,971 OSP and 1,511 PEG users. Relative use of OSP versus PEG decreased from 88.0% before 2004 to 48.4% in 2006. 70.2% of OSP users had no recorded creatinine determination within 60 days before colonoscopy, and this proportion did not decrease over time. The study population included 317 patients with a baseline GFR < 60 ml/min, and we identified one case with an unexplained creatinine increase > or = 0.5 mg/dl among 191 PEG users (0.5%) versus eight cases among 126 OSP users (6.3%). Unadjusted and adjusted relative risk estimates on comparing OSP with PEG were 12.1 (95% CI, 1.5-95.8) and 12.6 (95% CI, 1.5-106.5), respectively. CONCLUSIONS: In patients with preexisting renal disease, OSP use was associated with an increased risk of aggravated renal dysfunction versus PEG. Creatinine measurement with GFR estimation should be done before OSP administration in order to avoid its use in patients with renal disease.

Is there an association between hypothyroidism and open-angle glaucoma in an elderly population? An epidemiologic study.

PURPOSE: There have been conflicting reports pertaining to the association between hypothyroidism and open-angle glaucoma (OAG). The purpose of this study was to assess the hypothesized association between preexisting hypothyroidism and development of OAG in a population-based setting. DESIGN: Case-control study. PARTICIPANTS: The study population and controls were taken from all patients in a large US managed care database aged >or=60 years with 4 years of continuous eligibility dating from January 1, 2001, through December 31, 2004. METHODS: A total of 4728 newly diagnosed OAG patients were matched with 14 184 controls (3:1 matching) based on age and gender. MAIN OUTCOME MEASURES: Conditional logistic regression was used to assess the relationship between hypothyroidism and OAG while controlling for various risk factors (ischemic heart disease, cerebrovascular disease, hyperlipidemia, hypertension, arterial disease, diabetes, and migraines). RESULTS: Based on a diagnosis of hypothyroidism or use of a thyroid replacement therapy, prior hypothyroidism was found in 815 (17.2%) OAG subjects and in 2498 (17.6%) control subjects. After adjusting for the specified risk factors, patients with OAG were not found to be associated with a prior hypothyroid diagnosis when compared with control subjects (odds ratio, 0.93; 95% confidence interval, 0.85-1.01). CONCLUSIONS: An association between prior hypothyroidism and OAG development was not found. The large proportion of patients receiving thyroid replacement therapy may have negated any OAG-related consequences of hypothyroidism.

Risk of impaired renal function after colonoscopy: a cohort study in patients receiving either oral sodium phosphate or polyethylene glycol.

OBJECTIVES: We aimed to evaluate frequency, predictors, and monitoring of renal dysfunction related to the use of oral sodium phosphates for colonoscopy in clinical practice. METHODS: Cohort study using clinical records and electronic patient information from the Henry Ford Health System, Detroit, MI. We identified patients undergoing colonoscopy using sodium phosphate or polyethylene glycol (PEG), and estimated the risk of renal impairment associated with bowel preparation and other risk factors. RESULTS: Out of 7,897 patients, 6,833 had used sodium phosphate; 1,617 patients had renal dysfunction within 12 months prior to colonoscopy and 3,928 patients had no creatinine measurement within 12 months prior to or 6 months postcolonoscopy. Among the remaining 2,352 patients, 88 had incident renal dysfunction (glomerular filtration rate [GFR]<60 mL/min) after colonoscopy. The relative risk (RR) estimate for renal dysfunction comparing sodium phosphate with PEG was 1.13 (95% CI 0.58-2.23) without adjustment, and 1.14 (95% CI 0.55-2.39) after multivariate adjustment. Significant univariate risk factors were age>or=65 yr, African-American race, low baseline GFR, hypertension, and use of angiotensin-converting enzyme inhibitors (ACE inhibitors), angiotensin-renin blockers, or thiazide diuretics. CONCLUSIONS: In patients without preexisting renal disease, the risk of renal impairment after colonoscopy appears to be similar between sodium phosphate and PEG users. Sodium phosphate use in patients with preexisting renal disease is not recommended, but common in clinical practice. Sodium phosphate should not be used in patients with preexisting serious renal disease, adequate hydration should be assured in all patients, and renal function should be monitored before and after colonoscopy in those at risk of renal dysfunction.

Temporal relationship between use of NSAIDs, including selective COX-2 inhibitors, and cardiovascular risk.

BACKGROUND AND OBJECTIVE: The search for NSAIDs with less gastrointestinal toxicity led to the introduction of the selective cyclo-oxygenase-2 (COX-2) inhibitors. However, following their introduction into the market, concerns have developed regarding their safety, particularly their cardiovascular safety. The purpose of this study was to assess the cardiovascular risk (events included were myocardial infarction, stroke and myocardial infarction-related deaths) associated with long-term (>180 days of exposure) and short-term (or=35 years of age who received celecoxib, rofecoxib, ibuprofen, etodolac and naproxen from 1 January 1999 through 31 December 2001, were included. Multivariate Cox proportional hazard models were used to analyse the relationship between cardiovascular risk and NSAID use, including selective COX-2 inhibitor use, while adjusting for various risk factors. RESULTS: We identified 12 188 exposure periods (11 930 persons) and 146 cardiovascular events over the entire study period. Compared with long-term ibuprofen use, long-term use of celecoxib (adjusted hazard ratio [HR] 3.64; 95% CI 1.36, 9.70) and rofecoxib (adjusted HR 6.64; 95% CI 2.17, 20.28) was associated with a significant increase in cardiovascular risk. When restricted to patients >or=65 years of age, the cardiovascular risks associated with long-term celecoxib (adjusted HR 7.36; 95% CI 1.62, 33.48) and rofecoxib (adjusted HR 13.24; 95% CI 2.59, 67.68) use increased. Short-term use of celecoxib (adjusted HR 0.75; 95% CI 0.42, 1.35) and rofecoxib (adjusted HR 0.85; 95% CI 0.39, 1.86) was not associated with any significant change in cardiovascular risk when compared with short-term ibuprofen use. Neither long- nor short-term exposure to naproxen and etodolac was associated with cardionegative or cardioprotective effects when compared with ibuprofen use. CONCLUSIONS: The findings of this observational study, along with recent clinical trial results, suggest that prolonged exposure to selective COX-2 inhibitors may be associated with an increased risk of adverse cardiovascular outcomes.