Several observational studies have indicated that vitamin D receptor activators (VDRA), including paricalcitol, are associated with greater survival in maintenance hemodialysis (MHD) patients. However, patients with higher serum parathyroid hormone (PTH), a surrogate of higher death risk, are usually given higher VDRA doses, which can lead to confounding by indication and attenuate the expected survival advantage of high VDRA doses.
We examined mortality-predictability of low (>1 but <10 μg/week) versus high (≥10 μg/week) dose of administered paricalcitol over time in a contemporary cohort of 15,442 MHD patients (age 64±15 years, 55% men, 44% diabetes, 35% African Americans) from all DaVita dialysis clinics across the USA (7/2001-6/2006 with survival follow-ups until 6/2007) using conventional Cox regression, propensity score (PS) matching, and marginal structural model (MSM) analyses.
In our conventional Cox models and PS matching models, low dose of paricalcitol was not associated with mortality either in baseline (hazard ratio (HR): 1.03, 95%confidence interval (CI): (0.97-1.09)) and (HR: 0.99, 95%CI: (0.86-1.14)) or time-dependent (HR: 1.04, 95%CI: (0.98-1.10)) and (HR: 1.12, 95%CI: (0.98-1.28)) models, respectively. In contrast, compared to high dose of paricalcitol, low dose was associated with a 26% higher risk of mortality (HR: 1.26, 95%CI: (1.19-1.35)) in MSM. The association between dose of paricalcitol and mortality was robust in almost all subgroups of patients using MSMs.
Higher dose of paricalcitol appears causally associated with greater survival in MHD patients. Randomized controlled trials need to verify the survival effect of paricalcitol dose in maintenance hemodialysis patients are indicated.
Mortality; propensity score; marginal structural model; mineral and bone disorders; chronic kidney disease; paricalcitol
Significantly increasing heart transplantations have been performed in Taiwan in the past decades, but the trends of maintenance immunosuppression for heart transplant recipients have not been well known. In this study, we aimed to explore the trends of maintenance immunosuppressive therapy and common complications for heart transplant recipients.
We retrospectively analyzed ambulatory prescriptions in 488 heart transplant recipients for the period 2000–2009. Patient complications after heart transplantation were also identified.
The annual number of new heart transplant recipients ranged from 18 to 68. The 5-year survival rate was 77.9%. The total number of regimens was 10 in 2000, and increased to 28 in 2009. Most prescriptions were immunosuppressive combinations (95.5%–89.5%). The majority of immunosuppressive regimens were a triple regimen: cyclosporine, mycophenolic acid and corticosteroid in 2009. Cyclosporine was a predominant calcineurin inhibitor with a decreasing trend from 73.9% to 59.1%, whereas the use of tacrolimus significantly increased from 11.9% to 38.4%. Mycophenolic acid was the most frequently used antimetabolite (60.1%–80.3%), while the use of azathioprine was reduced (21.6%–2.3%). From 2008, the launch of everolimus initiated a new era in the utilization of mammalian target of rapamycin inhibitors for maintenance immunosuppression.
Cyclosporine remained the most frequently used calcineurin inhibitors, and tacrolimus increased gradually. Mycophenolic acid was the most popular antimetabolite rather than azathioprine. The rapidly increased everolimus combined regimen may change the patterns of maintenance immunosuppression. The increasing number of combination therapies indicates an active role of everolimus and a tendency of complex tailored individual therapies. © 2014 The Authors. Pharmacoepidemiology and Drug Safety published by John Wiley & Sons Ltd.
heart transplantation; immunosuppressive therapy; prescription patterns; pharmacoepidemiology
To evaluate the advantages and disadvantages of pre-approval requirements for safety data to detect cardiovascular (CV) risk contained in the December 2008 U.S. Food and Drug Administration (FDA) guidance for developing type 2 diabetes drugs compared with the February 2008 FDA draft guidance from the perspective of diabetes population health.
We applied the incremental net health benefit (INHB) framework to quantify the benefits and risks of investigational diabetes drugs using a common survival metric (life-years [LYs]). We constructed a decision analytic model for clinical program development consistent with the requirements of each guidance and simulated diabetes drugs, some of which had elevated CV risk. Assuming constant research budgets, we estimate the impact of increased trial size on drugs investigated. We aggregate treatment benefit and CV risks for each approved drug over a 35-year horizon under each guidance.
The quantitative analysis suggests that the December 2008 guidance adversely impacts diabetes population health. INHB was −1.80 million LYs, attributable to delayed access to diabetes therapies (−0.18 million LYs) and fewer drugs (−1.64 million LYs), but partially offset by reduced CV risk exposure (0.02 million LYs). Results were robust in sensitivity analyses.
The health outcomes impact of all potential benefits and risks should be evaluated in a common survival measure, including health gain from avoided adverse events, lost health benefits from delayed or forgone efficacious products, and impact of alternative policy approaches. Quantitative analysis of the December 2008 FDA guidance for diabetes therapies indicates that negative impact on patient health will result. © 2014 The Authors. Pharmacoepidemiology and Drug Safety published by John Wiley & Sons, Ltd.
INHB; benefit; risk; CV; evaluation; pharmacoepidemiology
To examine the age and gender-specific trends of schedule II opioid use among California residents, with special reference to multiple provider users (“doctor shoppers”).
Utilizing data from the California Prescription Drug Monitoring Program, we examined age and gender-specific trends of Schedule II opioid use during calendar years 1999–2007. Specifically, we analyzed: 1) the prevalence of Schedule II opioid users among California’s population, and 2) the proportion of these opioid users who were doctor shoppers (defined as an individual who used more than five different prescribers for all schedule II opioids he/she obtained in a calendar year).
Among all age and gender groups, the prevalence of Schedule II opioid users in California increased by 150%–280% and the prevalence of doctor shoppers among users increased by 111%–213% over nine years. The prevalence of opioid users was lowest among 18–44 year-old males (1.25%) and highest among 65 years and older females (5.31%) by 2007. The prevalence of doctor shoppers was approximately 1.4% among those up to age 64 and 0.5% among those 65 years and older. The gender difference in doctor shoppers among all age groups was negligible. On average, the cumulative morphine-equivalent amount of Schedule II opioid per individual obtained per year was three- to six-fold higher for doctor shoppers than for the general population across different age and gender groups.
Age and gender differences in opioid use were relatively small, while the trends for use of opioids and multiple providers grew at a disquieting rate.
Age; gender; opioid prescription; doctor shopping; prevalence; California
The Infectious Diseases Society of America and US CDC recommend 60 days of ciprofloxacin, doxycycline or amoxicillin for anthrax prophylaxis. It is not possible to determine severe adverse drug event (ADE) risks from the few people thus far exposed to anthrax prophylaxis. This study’s objective was to estimate risks of severe ADEs associated with long-term ciprofloxacin, doxycycline and amoxicillin exposure using 3 large databases: one electronic medical record (General Practice Research Database) and two claims databases (UnitedHealthcare, HMO Research Network).
We include office visit, hospital admission and prescription data for 1/1/1999–6/30/2001. Exposure variable was oral antibiotic person-days (pds). Primary outcome was hospitalization during exposure with ADE diagnoses: anaphylaxis, phototoxicity, hepatotoxicity, nephrotoxicity, seizures, ventricular arrhythmia or infectious colitis.
We randomly sampled 999,773, 1,047,496 and 1,819,004 patients from Databases A, B and C respectively. 33,183 amoxicillin, 15,250 ciprofloxacin and 50,171 doxycycline prescriptions continued ≥30 days. ADE hospitalizations during long-term exposure were not observed in Database A. ADEs during long-term amoxicillin were seen only in Database C with 5 ADEs or 1.2(0.4–2.7) ADEs/100,000 pds exposure. Long-term ciprofloxacin showed 3 and 4 ADEs with 5.7(1.2–16.6) and 3.5(1.0–9.0) ADEs/100,000 pds in Databases B and C, respectively. Only Database B had ADEs during long-term doxycycline with 3 ADEs or 0.9(0.2–2.6) ADEs/100,000 pds. For most events, the incidence rate ratio, comparing >28 vs.1–28 pds exposure was <1, showing limited evidence for cumulative dose-related ADEs from long-term exposure.
Long-term amoxicillin, ciprofloxacin and doxycycline appears safe, supporting use of these medications if needed for large-scale post-exposure anthrax prophylaxis.
Anti-infective agents; Anthrax prevention and control; Health services research; Databases; Bioterrorism
Use of administrative or population-based databases for post-marketing pharmacoepidemiology research in patients with end-stage liver disease (ESLD) has been limited by the difficulty of accurately identifying such patients. Algorithms to identify patients with ESLD using ICD-9-CM codes have not been developed outside of the Veterans Affairs healthcare setting.
We queried electronic medical records at two tertiary care hospitals to identify patients with ICD-9-CM codes indicative of ESLD. Coding algorithms were developed to identify patients with confirmed ESLD, and these were tested to determine their positive predictive value (PPV).
The presence of one inpatient or outpatient ICD-9-CM code for: a) cirrhosis, b) chronic liver disease, and c) a hepatic decompensation event yielded a PPV of 85.2% (167/196; 95% CI: 79.4%–89.9%). The PPV increased to 89.3% (150/168; 95% CI: 83.6%–93.5%) when the algorithm required 2 or more ICD-9-CM codes for a hepatic decompensation. However, an algorithm requiring only one ICD-9-CM code for a) cirrhosis and b) a hepatic decompensation event, in the absence of a chronic liver disease code, yielded a PPV of 85.7% (30/35; 95% CI: 69.7%–95.2%).
A coding algorithm that includes at least one ICD-9-CM code for cirrhosis plus one ICD-9-CM code for a hepatic decompensation event has a high PPV for identifying patients with ESLD. The inclusion of at least 2 codes indicative of chronic liver disease increased the PPV. This algorithm can be used in future epidemiologic studies to examine the outcomes of a variety of long-term medical therapies in patients with ESLD.
Validation; hepatic decompensation; cirrhosis; ICD-9-CM code; end-stage liver disease
HIV-infected patients may be at particular risk for acetaminophen-induced hepatotoxicity, but acetaminophen use in the context of liver injury has been incompletely examined among HIV-infected patients. Among a sample of HIV-infected patients, we aimed to determine acetaminophen exposure; assess the cross-sectional association between acetaminophen exposure and advanced hepatic fibrosis; and determine whether factors associated with acetaminophen exposure varied by HCV status.
We conducted a cross-sectional analysis of the Veterans Aging Cohort Study. Advanced hepatic fibrosis was defined as a FIB-4 > 3.25, a composite score calculated based on age, alanine aminotransferase, aspartate aminotransferase and platelet count. Multivariable ordered polytomous logistic regression was used to determine the association between FIB-4 status and acetaminophen exposure stratified by HCV status.
Among HIV-infected patients (n=14,885), 31% received at least one acetaminophen prescription. Among those receiving acetaminophen, acetaminophen overuse was common among both HIV-monoinfected and HIV/HCV-coinfected patients (846 [31%] vs. 596[32%], p=0.79). After stratifying by HCV status, those with evidence of advanced liver fibrosis were equally likely to be exposed to acetaminophen. Further, HIV-monoinfected patients with an alcohol use disorder were more likely to have acetaminophen overuse (OR [95% CI]=1.56 [1.21, 2.02]).
Strategies to minimize acetaminophen exposure, especially for HIV-monoinfected patients, are warranted.
acetaminophen; HIV; Hepatitis C; Veterans; medication
To describe the rationale and methods for a prospective, open-cohort study assessing the long-term safety of Prolia® for treatment of postmenopausal osteoporosis (PMO) in postmarketing settings.
Data will be derived from United States Medicare, United Healthcare, and Nordic (Denmark, Sweden, Norway) national registries. Observation will begin on the date of first Prolia® regulatory approval (May 26, 2010) and continue for 10 years. Women with PMO will be identified by postmenopausal age, osteoporosis diagnosis, osteoporotic fracture, or osteoporosis treatment. Exposure to Prolia® and bisphosphonates will be updated during follow-up; exposure cohorts will be defined based on patient-years during which patients are on- or post-treatment. Nine adverse events (AEs) will be assessed based on diagnosis codes: osteonecrosis of the jaw (ONJ), atypical femoral fracture (AFF), fracture healing complications, hypocalcemia, infection, dermatologic AEs, acute pancreatitis, hypersensitivity, and new primary malignancy. Medical review will confirm selected potential cases of ONJ and AFF. Incidence rates (IRs) of AEs will be described overall and for exposure cohorts; multivariate Cox proportional hazard regression models will compare IRs of AEs across exposure cohorts. Utilization patterns of Prolia® for approved, and unapproved indications will be described.
This study is based on comprehensive preliminary research and considers methodological challenges specific to the study population. The integrated data systems used in this regulatory committed program can serve as a powerful data resource to assess diverse and rare AEs over time. © 2013 Amgen Inc. Pharmacoepidemiology and Drug Safety published by John Wiley & Sons, Ltd.
postmenopausal osteoporosis; Prolia® (denosumab); postmarketing drug safety; pharmacovigilance; database; pharmacoepidemiology methods; pharmacoepidemiology
The purpose of this study was to investigate the impact of risk communication about bisphosphonate (BP)-related osteonecrosis of the jaw (ONJ) on the number of reported cases to the Drug Adverse Reactions Reporting System and on the incidence proportion of ONJ in a hospital-based cohort study in Japan.
We conducted a survey of the safety information on BP-related ONJ available from regulatory authorities, pharmaceutical manufacturers and academic associations. We also performed a trend analysis of a dataset from the Drug Adverse Reactions Reporting System and a sub-analysis, using previously constructed data from a retrospective cohort study.
Risk communication from pharmaceutical manufacturers and academic associations began within 1 year after revisions were made to the package inserts, in October 2006. Twenty times more cases of ONJ have been reported to regulatory authority since 2007, compared with the period before 2007. In our cohort, the incidence proportion of ONJ during and after 2009 was four times greater than before 2009. During this period, BPs were frequently prescribed, whereas there was no increase in the use of alternative agents, such as selective estrogen receptor modulators.
ONJ was increasingly diagnosed after risk communication efforts, but the impact of the communications was not clear. Safety notifications were diligently disseminated after the package insert was revised. However, there was no surveillance for ONJ before the revision. © 2014 The Authors. Pharmacoepidemiology and Drug Safety published by John Wiley & Sons, Ltd.
risk communication; osteonecrosis of the jaw; oral bisphosphonates; pharmacoepidemiology
The purpose of this study is to further evaluate the safety of the human papillomavirus (HPV)-16/18-AS04-adjuvanted vaccine (HPV-16/18-vaccine Cervarix®, GlaxoSmithKline, Belgium) through a pooled analysis of data from 42 completed/ongoing clinical studies.
Unsolicited adverse events (AEs) were reported for 30 days after each dose. Medically significant conditions, serious AEs (SAEs), potential immune-mediated diseases (pIMDs) and pregnancy outcomes were captured until study completion. Events leading to subject withdrawal were reviewed. Relative risks compared incidences of spontaneous abortion and pIMDs in controlled studies.
Thirty one thousand one hundred seventy-three adolescent girls/women received HPV-16/18-vaccine alone (HPV group), 2166 received HPV-16/18-vaccine coadministered with another vaccine and 24 241 were controls. Mean follow-up was 39 months (range 0–113.3). Incidences of unsolicited AEs reported within 30 days after any dose were similar between HPV and Control groups (30.8%/29.7%). During the entire study period, reports of medically significant conditions (25.0%/28.3%) and SAEs (7.9%/9.3%) were also similarly distributed between groups. Deaths were rare: HPV (alone/coadministered) n = 25, controls n = 20 (n = 18 in blinded groups). pIMDs within 1 year were reported by 0.2% of HPV-16/18 vaccinees and controls. For each pIMD event category, no increased relative risks were reported for HPV-16/18 vaccinees versus controls. Coadministration did not change the overall safety profile. Pregnancy outcomes and withdrawal rates were similar between groups.
Analysis of safety data arising from 57 580 subjects and 96 704 HPV-16/18-vaccine doses shows that the incidences and distribution of AEs were similar among HPV-16/18-vaccine recipients and controls. No new safety signals were identified. The data confirm previous findings that HPV-16/18-vaccine has an acceptable benefit-risk profile in adolescent girls and adult women.
human papillomavirus vaccine; safety; adverse drug reactions; pregnancy; autoimmune disease; AS04; pharmacoepidemiology
To obtain safety and effectiveness data on a combined anti-HIV drug, Epzicom (abacavir 600 mg/lamivudine 300 mg), a post-marketing surveillance on Epzicom that was required by the Japanese regulatory authority was conducted between January 2005 and December 2010.
A joint survey (HIV-related drug [HRD] survey) has been conducted involving manufacturers of drugs for treatment of HIV infection in Japan. Safety and effectiveness data from total 624 cases (1107.3 person-years) registered to the HRD surveys and received Epzicom were obtained. Adverse drug reactions (ADRs) were defined as adverse events (AE) of which association with Epzicom could not be ‘ruled out’.
It was found that the incidence of ADR was 32.4% (202/624 cases) on the case basis. In addition, the frequently reported ADR included hyperlipidaemia (59 cases), hypertriglyceridaemia (21 cases), blood bilirubin increased (19 cases), gamma-glutamyltransferase increase (14 cases), blood triglyceride increase (14 cases) and rash (14 cases). Serious AEs were seen in 19 patients (30 events), including one death (no evident association with Epzicom). There were four cases (0.6%) of survey-defined ‘hypersensitivity’, and the incidence was 0.9% (4/445) among abacavir naïve patients; none of which was reported as serious. No case of myocardial infarction was reported. One pregnant case who delivered a normal baby by caesarean section was reported to have experienced aggravation of anaemia and nausea.
The post-marketing surveillance indicated that the incidence of both ischaemic heart disease and hypersensitivity associated with Epzicom was considerably low, suggesting that this drug can be safely used in the Japanese population.
abacavir; lamivudine; human immunodeficiency virus; hypersensitivity; myocardial infarction; Japan; pharmacoepidemiology
Determining the nature of the relationship between cumulative duration of exposure to an agent and the hazard of an adverse outcome is an important issue in environmental and occupational epidemiology, public health and clinical medicine. The Cox proportional hazards regression model can incorporate time-dependent covariates. An important class of continuous time-dependent covariates is that denoting cumulative duration of exposure.
We used fractional polynomial methods to describe the association between cumulative duration of exposure and adverse outcomes. We applied these methods in a cohort study to examine the relationship between cumulative duration of use of the antiarrhythmic drug amiodarone and the risk of thyroid dysfunction. We also used these methods with a conditional logistic regression model in a nested case-control study to examine the relationship between cumulative duration of use of bisphosphonate medication and the risk of atypical femur fracture.
Using a cohort design and a Cox proportional hazards model, we found a non-linear relationship between cumulative duration of use of the antiarrhythmic drug amiodarone and the risk of thyroid dysfunction. The risk initially increased rapidly with increasing cumulative use. However, as cumulative duration of use increased, the rate of increase in risk attenuated and eventually levelled off. Using a nested case-control design and a conditional logistic regression model, we found evidence of a linear relationship between duration of use of bisphosphonate medication and risk of atypical femur fractures.
Fractional polynomials allow one to model the relationship between cumulative duration of medication use and adverse outcomes.
Cox proportional hazards regression model; fractional polynomials; time-dependent covariates; survival analysis; pharmacoepidemiology; cohort study; nested case-control study
Abacavir is a nucleoside reverse transcriptase inhibitor indicated for human immunodeficiency virus (HIV) infection. In Japan, Ziagen® (300-mg abacavir sulfate) has been marketed since 1999. To obtain safety data on Ziagen, a mandatory postmarketing surveillance was conducted between September 1999 and September 2009.
A joint survey [HIV-related Drug Surveys (HRD)] has been conducted involving manufacturers of drugs for HIV treatment in Japan. Safety data from total 643 cases (1345.7 person-years) registered to the HRD surveys and received Ziagen were obtained. Adverse drug reaction (ADR) was defined as adverse event of which association with abacavir could not be “ruled out.”
It was found that the overall frequency of ADR was 47.6% (306/643); the common ADRs were “hyperlipidemia,” “nausea,” “increased γ-glutamyltransferase level,” “increased blood triglycerides,” “abnormal hepatic function,” and so on. Serious adverse events were reported in 65 subjects; however, none of the three fatal cases were clearly associated with Ziagen use. The survey-defined hypersensitivity has been infrequently reported in 15 subjects (2.3%). Although some studies had indicated of the association between abacavir and myocardial infarction, no ischemic heart diseases were reported in the present survey. Two of the three pregnant cases delivered normal neonates (one induced abortion).
During the mandatory postmarketing survey of Ziagen, there were no cases of ischemic heart diseases, and the incidence of hypersensitivity was considerably low. These indicated that abacavir can be safely used in Japanese HIV+ population. However, the safety profile of Ziagen should be continued to be monitored through pharmacovigilance.
abacavir; human immunodeficiency virus; hypersensitivity; myocardial infarction; adverse drug reactions; Japanese; pharmacoepidimiology
After U.S. licensure, parenterally administered medications are identified using non-specific drug codes. Accurately identifying these medications is critical to safety and effectiveness research. Methods to identify medications prior to assignment of specific drug codes have not been well described.
To describe a generalized approach using non-specific drug codes to identify parenteral therapies in Medicare claims and to assess the ability of that approach to identify tocilizumab (TCZ), a new biologic agent approved in 2010.
We used 2008–2010 Medicare data for a cohort of rheumatoid arthritis patients for algorithm development. Our algorithm classified non-specific drug codes based upon: (1) ICD9 codes; (2) unit values (i.e. dose); (3) codes for infusion/injection procedures; (4) expected versus observed total reimbursement amount and reimbursement per unit. We assessed algorithm performance by linking to an arthritis registry to examine external validity.
Of 472,803 claims with non-specific drug codes, 9,762 claims satisfied the TCZ algorithm. 74.3% of 9,762 claims were classified as TCZ by exact unit price or allowed amount, 4.4% by unique doses, 21.3% by diagnosis code and small deviation from unit price or allowed amount. The algorithm demonstrated good performance characteristics: sensitivity 94% (95% CI 80–99), specificity 100% (99–100) and PPV 97% (84–100).
Claims-based algorithms in Medicare or similar data systems can accurately identify newly approved biologics administered parenterally prior to the assignment of specific drug codes.
rheumatoid arthritis; Medicare; Part D; biologics; tocilizumab; denosumab; certolizumab; linkage; registry
High-grade cervical dysplasia or cervical intraepithelial neoplasia (CIN) grade 2 or worse has been widely used as a surrogate endpoint in cervical cancer screening or prevention trials.
To identify high-grade cervical dysplasia and cervical cancer, we developed claims-based algorithms that incorporated a combination of diagnosis and procedure codes using the billing data in an electronic medical records (EMR) database and assessed the validity of the algorithms in an independent administrative claims database. We calculated the positive predictive value (PPV) with the 95% confidence interval (CI) of each algorithm, using new cytologic or pathologic diagnosis of CIN 2 or 3, carcinoma in situ, or cervical cancer as the gold standard.
Having ≥1 diagnosis code for high-grade cervical dysplasia or cervical cancer had a PPV of 57.1% (95%CI 54.7–59.5%). By requiring ≥2 diagnoses for high-grade cervical dysplasia or cervical cancer, separated by 7 to 30 days, the PPV increased to 60.2% (95%CI 53.9–66.1%). At least 2 diagnoses and a procedure code within a month from the first diagnosis date yielded a PPV of 80.7% (95%CI 73.6–86.2%). The algorithms had greater PPVs in identifying prevalent high-grade cervical dysplasia or cervical cancer. Overall, the PPVs of these algorithms were similar or slightly lower in the external claims data than in the sample used to derive the algorithms.
Use of ≥ 2 diagnosis codes in combination with a procedure code appears to be a valid tool for studying high-grade cervical dysplasia and cervical cancer in both EMR and administrative claims databases.
cervical dysplasia; cervical cancer; validity; claims data
To develop and validate the positive predictive value (PPV) of an algorithm to identify anaphylaxis using health plan administrative and claims data. Previously published positive predictive values (PPVs) for anaphylaxis using ICD-9-CM codes range from 52-57%.
We conducted a retrospective study using administrative and claims data from eight health plans. Using diagnosis and procedure codes, we developed an algorithm to identify potential cases of anaphylaxis from the Mini-Sentinel Distributed Database between January 2009 and December 2010. A random sample of medical charts (N=150) was identified for chart abstraction. Two physician adjudicators reviewed each potential case. Using physician adjudicator judgments on whether the case met diagnostic criteria for anaphylaxis, we calculated a PPV for the algorithm.
Of the 122 patients for whom complete charts were received, 77 were judged by physician adjudicators to have anaphylaxis. The PPV for the algorithm was 63.1% (95% CI: 53.9%-71.7%), using the clinical criteria by Sampson as the gold standard. The PPV was highest for inpatient encounters with ICD-9-CM codes of 995.0 or 999.4. By combining only the top performing ICD-9-CM codes, we identified an algorithm with a PPV of 75.0%, but only 66% of cases of anaphylaxis were identified using this modified algorithm.
The PPV for the ICD-9-CM-based algorithm for anaphylaxis was slightly higher than PPV estimates reported in prior studies, but remained low. We were able to identify an algorithm which optimized the PPV but demonstrated lower sensitivity for anaphylactic events.
anaphylaxis; serious allergic reaction; validation; administrative data; Food and Drug administration; Mini-Sentinel
This paper addresses strategies for selecting variables for adjustment in non-experimental comparative effectiveness research (CER), and uses causal graphs to illustrate the causal network that relates treatment to outcome. Variables in the causal network take on multiple structural forms. Adjustment for on a common cause pathway between treatment and outcome can remove confounding, while adjustment for other structural types may increase bias. For this reason variable selection would ideally be based on an understanding of the causal network; however, the true causal network is rarely know. Therefore, we describe more practical variable selection approaches based on background knowledge when the causal structure is only partially known. These approaches include adjustment for all observed pretreatment variables thought to have some connection to the outcome, all known risk factors for the outcome, and all direct causes of the treatment or the outcome. Empirical approaches, such as forward and backward selection and automatic high-dimensional proxy adjustment, are also discussed. As there is a continuum between knowing and not knowing the causal, structural relations of variables, we recommend addressing variable selection in a practical way that involves a combination of background knowledge and empirical selection and that uses the high-dimensional approaches. This empirical approach can be used to select from a set of a priori variables based on the researcher’s knowledge to be included in the final analysis or to identify additional variables for consideration. This more limited use of empirically-derived variables may reduce confounding while simultaneously reducing the risk of including variables that may increase bias.
covariate selection; confounding; comparative effectiveness research; propensity scores; backwards selection; stepwise selection; bias; nonexperimental methods
Statins are widely used for preventing cardiovascular disease, yet recent reports suggest an increased risk of acute kidney injury (AKI). We estimated the one-year risk of AKI associated with statin initiation and determined the comparative safety of individual statin formulations.
We performed a cohort study in insurance billing data from commercial and Medicare insurance plans in the United States for the years 2000—2010. We identified statin initiators and non-users with histories of medication use and healthcare utilization. AKI diagnosis codes were identified in the one year following the index date. We estimated hazard ratios (HR) and 95% confidence intervals (CI) with adjusted and propensity score (PS)-matched Cox-proportional hazards models. Models were run separately in insurance groups and adjusted for cardiovascular and renal risk factors, markers of healthcare utilization, and other medication use.
We identified 3,905,155 statin initiators and 2,817,621 eligible non-users. The adjusted HR of AKI in statin initiators compared to non-users was: commercial, HR=1.04 (95% CI: 0.99, 1.09); Medicare, HR=0.72 (95% CI: 0.70, 0.75). PS-matching yielded: commercial, HR=0.82 (95% CI: 0.78, 0.87); Medicare HR=0.66 (95% CI: 0.63, 0.69). As individual formulations, higher-potency simvastatin was associated with an increased risk of AKI over lower-potency simvastatin in adjusted models: commercial, HR=1.42 (95% CI: 1.28, 1.58Medicare, HR=1.24 (95% CI: 1.15, 1.35).
As a class, statin initiation was not associated with an increase in AKI. However, higher-potency simvastatin did exhibit an increased AKI risk.
comparative effectiveness; drug safety; acute kidney injury; propensity scores
To conduct a synthesis of the literature on methods to evaluate the impacts of FDA regulatory actions, and identify best practices for future evaluations.
We searched MEDLINE for manuscripts published between January 1948 and August 2011 that included terms related to FDA, regulatory actions, and empirical evaluation; the review additionally included FDA-identified literature. We used a modified Delphi method to identify preferred methodologies. We included studies with explicit methods to address threats to validity, and identified designs and analytic methods with strong internal validity that have been applied to other policy evaluations.
We included 18 studies out of 243 abstracts and papers screened. Overall, analytic rigor in prior evaluations of FDA regulatory actions varied considerably; less than a quarter of studies (22%) included control groups. Only 56% assessed changes in the use of substitute products/services, and 11% examined patient health outcomes. Among studies meeting minimal criteria of rigor, 50% found no impact or weak/modest impacts of FDA actions and 33% detected unintended consequences. Among those studies finding significant intended effects of FDA actions, all cited the importance of intensive communication efforts. There are preferred methods with strong internal validity that have yet to be applied to evaluations of FDA regulatory actions.
Rigorous evaluations of the impact of FDA regulatory actions have been limited and infrequent. Several methods with strong internal validity are available to improve trustworthiness of future evaluations of FDA policies.
FDA; Regulatory Actions; Evaluation Methodology
Large, observational HIV cohorts play an important role in answering questions which are difficult to study in randomized trials; however, they often lack detailed information regarding previous antiretroviral treatment (ART). Knowledge of ART treatment history is important when ascertaining the long-term impact of medications, co-morbidities, or adverse reactions on HIV outcomes.
We performed a retrospective study to validate a prediction algorithm for identifying ART-naïve patients using the Veterans Aging Cohort Study’s Virtual Cohort—an observational cohort of 40,594 HIV-infected veterans nationwide. Medical records for 3070 HIV-infected patients were reviewed to determine history of combination ART treatment. An algorithm using Virtual Cohort laboratory data was used to predict ART treatment status and compared to medical record review.
Among 3070 patients’ medical records reviewed, 1223 were eligible for analysis. Of these, 990 (81%) were ART naïve at cohort entry based on medical record review. The prediction algorithm’s sensitivity was 86%, specificity 47%, positive predictive value (PPV) 87% and negative predictive value 45%, using a viral load threshold of <400 copies/ml. Sensitivity analysis revealed that PPV would be maximized by increasing the viral load threshold, whereas sensitivity would be maximized by lowering the viral load threshold.
A prediction algorithm using available laboratory data can be used to accurately identify ART-naïve patients in large, observational HIV cohorts. Use of this algorithm will allow investigators to accurately limit analyses to ART-naïve patients when studying the contribution of ART to outcomes and adverse events.
HIV-1; Antiretroviral Therapy, Highly Active; Cohort Studies; HIV infections/epidemiology; HIV infections/drug therapy
To determine which medications are most commonly used by women in the first trimester of pregnancy and identify the critical gaps in information about fetal risk for those medications.
Self-reported first-trimester medication use was assessed among women delivering liveborn infants without birth defects and serving as control-mothers in two large case-control studies of major birth defects. The Teratology Information System (TERIS) expert Advisory Board ratings of quality and quantity of data available to assess fetal risk were reviewed to identify information gaps.
Responses from 5,381 mothers identified 54 different medication components used in the first trimester by at least 0.5% of pregnant women, including 31 prescription and 23 over-the-counter medications. The most commonly used prescription medication components reported were progestins from oral contraceptives, amoxicillin, progesterone, albuterol, promethazine, and estrogenic compounds. The most commonly used over-the-counter medication components reported were acetaminophen, ibuprofen, docusate, pseudoephedrine, aspirin, and naproxen. Among the 54 most commonly used medications, only two had ‘Good to Excellent’ data available to assess teratogenic risk in humans, based on the TERIS review.
For most medications commonly used in pregnancy, there are insufficient data available to characterize the fetal risk fully, limiting the opportunity for informed clinical decisions about the best management of acute and chronic disorders during pregnancy. Future research efforts should be directed at these critical knowledge gaps.
medication; pregnancy; fetal risk; teratogen
Nonexperimental studies of treatment effectiveness provide an important complement to randomized trials by including heterogeneous populations. Propensity scores (PS) are common in these studies, but may not adequately capture changes in channeling experienced by innovative treatments. We use calendar time-specific (CTS) PSs to examine the effect of oxaliplatin during dissemination from off-label to widespread use.
Stage III colon cancer patients aged 65+ initiating chemotherapy between 2003–06 were examined using cancer registry data linked with Medicare claims. Two PS approaches for receipt of oxaliplatin vs. 5-flourouricil were constructed using logistic models with key components of age, sex, substage, grade, census level income, and comorbidities: 1) a conventional, year-adjusted PS and 2) a CTS PS constructed and matched separately within 1-year intervals, then combined. We compared PS-matched hazard ratios (HR) for mortality using Cox models.
Oxaliplatin use increased significantly; 8%(n=86) of patients received it in the first time period vs. 52%(n=386) in the last. Channeling by comorbidities, income, and age appeared to change over time. The CTS PS improved covariate balance within calendar time strata and yielded an attenuated estimated benefit of oxaliplatin (HR=0.75) compared with the conventional PS (HR=0.69).
In settings where prescribing patterns have changed and calendar time acts as a confounder, a CTS PS can characterize changes in treatment choices and estimating separate PSs within specific calendar time periods may result in enhanced confounding control. To increase validity of CER, researchers should carefully consider drug lifecycles and effects of innovative treatment dissemination over time.