Chun R. Chao

Sociodemographic characteristics of members of a large, integrated health care system: comparison with US Census Bureau data.

BACKGROUNDnData from the memberships of large, integrated health care systems can be valuable for clinical, epidemiologic, and health services research, but a potential selection bias may threaten the inference to the population of interest.nnnMETHODSnWe reviewed administrative records of members of Kaiser Permanente Southern California (KPSC) in 2000 and 2010, and we compared their sociodemographic characteristics with those of the underlying population in the coverage area on the basis of US Census Bureau data.nnnRESULTSnWe identified 3,328,579 KPSC members in 2000 and 3,357,959 KPSC members in 2010, representing approximately 16% of the population in the coverage area. The distribution of sex and age of KPSC members appeared to be similar to the census reference population in 2000 and 2010 except with a slightly higher proportion of 40 to 64 year olds. The proportion of Hispanics/Latinos was comparable between KPSC and the census reference population (37.5% vs 38.2%, respectively, in 2000 and 45.2% vs 43.3% in 2010). However, KPSC members included more blacks (14.9% vs 7.0% in 2000 and 10.8% vs 6.5% in 2010). Neighborhood educational levels and neighborhood household incomes were generally similar between KPSC members and the census reference population, but with a marginal underrepresentation of individuals with extremely low income and high education.nnnCONCLUSIONSnThe membership of KPSC reflects the socioeconomic diversity of the Southern California census population, suggesting that findings from this setting may provide valid inference for clinical, epidemiologic, and health services research.

Identifying primary and recurrent cancers using a SAS-based natural language processing algorithm

Objective Significant limitations exist in the timely and complete identification of primary and recurrent cancers for clinical and epidemiologic research. A SAS-based coding, extraction, and nomenclature tool (SCENT) was developed to address this problem. Materials and methods SCENT employs hierarchical classification rules to identify and extract information from electronic pathology reports. Reports are analyzed and coded using a dictionary of clinical concepts and associated SNOMED codes. To assess the accuracy of SCENT, validation was conducted using manual review of pathology reports from a random sample of 400 breast and 400 prostate cancer patients diagnosed at Kaiser Permanente Southern California. Trained abstractors classified the malignancy status of each report. Results Classifications of SCENT were highly concordant with those of abstractors, achieving κ of 0.96 and 0.95 in the breast and prostate cancer groups, respectively. SCENT identified 51 of 54 new primary and 60 of 61 recurrent cancer cases across both groups, with only three false positives in 792 true benign cases. Measures of sensitivity, specificity, positive predictive value, and negative predictive value exceeded 94% in both cancer groups. Discussion Favorable validation results suggest that SCENT can be used to identify, extract, and code information from pathology report text. Consequently, SCENT has wide applicability in research and clinical care. Further assessment will be needed to validate performance with other clinical text sources, particularly those with greater linguistic variability. Conclusion SCENT is proof of concept for SAS-based natural language processing applications that can be easily shared between institutions and used to support clinical and epidemiologic research.

Changes in serum prostate-specific antigen levels and the identification of prostate cancer in a large managed care population.

The use of a single, elevated PSA level to screen for prostate cancer is controversial given its reported low specificity and the questionable benefits of PSA screening on prostate cancer mortality. Current guidelines in the USA recommend against screening using a single PSA measurement. Previous studies suggest that using changes in PSA level over time, or PSA velocity, may improve the detection of prostate cancer and/or aggressive disease; however, this is also controversial as other studies suggest PSA velocity does not improve detection and may further contribute to the overdetection of indolent prostate cancer. Given the questions that remain regarding the use of the rate of change in PSA as a screening tool for prostate cancer because of previous conflicting studies, which to date have included small, highly selected populations, this study adds to the existing knowledge by assessing this question in general practice settings among a large, diverse population.

An unmasking phenomenon in an observational post-licensure safety study of adolescent girls and young women.

Our recent experience in a post-licensure safety study of autoimmune conditions following the quadrivalent human papillomavirus vaccine in 189,629 girls and young women ages 9-26 years led us to question the adequacy of the exclusion of Day 0 events to prevent the erroneous association of prevalent conditions with vaccination. Of the 18 confirmed cases of Graves disease diagnosed in days 1-60 following vaccination, only 6 cases appeared to be truly new onset. Among the remaining 12 cases, 2 cases had abnormal thyroid stimulating hormone or thyroxine labs drawn prior to or on Day 0 but had no documented pre-existing symptoms. The other 10 cases had mention of symptoms of hyperthyroidism referencing a period prior to first HPV-4 dose. This unmasking phenomenon, due to health care visits that include vaccination and new workups of preexisting symptoms, may not be adequately controlled through the exclusion of Day 0 events.

Risk factors of herpes zoster among children immunized with varicella vaccine: results from a nested case-control study.

Background: Previous studies of varicella-zoster virus reactivation in children have provided little information on potential risk factors. The aim of this study was to investigate the effects of race, chronic medical conditions and treatments, and recent vaccination, on the risk of herpes zoster (HZ) in children vaccinated with one dose of varicella vaccine. Methods: Case subjects were identified from a cohort of subjects who were members of the Southern California Kaiser Permanente Health Plan and received primary immunization with a single-antigen live varicella vaccine at age ≤12 years from 2002 to 2008. Control subjects free of HZ during the study period were matched at a 5:1 ratio to each case subject on date of birth and sex. Race information was obtained from membership files, health records, and phone interview. Immunization history, medical history, and health care utilization were identified from Southern California Kaiser Permanente Health Plan electronic records. Results: During this time, 122 children were diagnosed with HZ. With adjustment for the number of hospitalizations, outpatient visits, and length of time between vaccination with varicella vaccine and the onset of HZ, Black children were at lower risk of developing HZ than were White (OR=0.41, 95% CI=0.17–0.98) and Asian children (OR=0.30, 95% CI=0.11–0.84). Conclusions: These data suggest that the racial differences in the risk of developing HZ seen in adults are manifest in children as well. As children are not subject to the majority of factors hypothesized to underlie HZ in adults and as this study was conducted in a setting which affords equal access to health care, it is possible that genetic variation may explain some portion of varicella-zoster virus reactivation.

The Accuracy and Trends of Smoking History Documentation in Electronic Medical Records in a Large Managed Care Organization

The accuracy of smoking history documentation in the electronic medical records was examined at a large managed care organization among 36,494 male members who self-reported smoking history in mailed surveys. The sensitivity of electronic smoking history documentation for ever-smoking status was 0.19 in years 2003–2005 (using ICD-9/CPT code only), 0.80 in 2006–2008 and 0.84 in 2009–2010 (combination of ICD-9/CPT codes and risk factor module used after 2006). The positive predictive value was 0.96, 0.90, and 0.95 in these periods, respectively. Among self-reported ever-smokers, increased healthcare utilization and smoking intensity/duration were associated with higher likelihood of having electronic smoking history documentation, while Asian race and Spanish language preference were associated with lower likelihood. These data suggest that enhanced efforts may be needed to screen for and document smoking among racial/ethnic minorities.

Quadrivalent Human Papillomavirus Vaccine Initiation in Boys Before and Since Routine Use: Southern California, 2009–2013

OBJECTIVESnWe examined the trends and correlates of quadrivalent human papillomavirus vaccine (HPV4) initiation in insured boys during the periods before and after routine use recommendation.nnnMETHODSnWe grouped data from electronic medical records of boys aged 9 to 17 years from the Kaiser Permanente Southern California prepaid health plan into 3 open cohorts: permissive use: 2009 to 2010; anal cancer indication added: 2010 to 2011; and routine use: 2011 to 2013. We estimated adjusted risk ratios (ARRs) between demographics and vaccination initiation using Poisson regression.nnnRESULTSnHPV4 initiation increased across cohorts--1.6%, 3.4%, and 18.5%--with the greatest increase among boys aged 11 to 12 years in cohort 3. Initiation was associated with receiving influenza vaccination in the previous year in all cohorts (cohort 3: ARRu2009=u20091.48; 95% confidence interval [CI]u2009=u20091.46, 1.51) and with non-White race/ethnicity following routine recommendation (cohort 3, non-Hispanic Black: ARRu2009=u20091.18; 95% CIu2009=u20091.08, 1.30; Hispanic: ARRu2009=u20091.23; 95% CIu2009=u20091.17, 1.29; Asian/Pacific Islanders: ARRu2009=u20091.16; 95% CIu2009=u20091.11, 1.20).nnnCONCLUSIONSnRoutine use recommendation increased the uptake of HPV4 in boys. System-level interventions to encourage providers to routinely recommend HPV4 vaccination may help increase HPV4 uptake in boys.

Unmasking in an observational vaccine safety study: Using type 2 diabetes mellitus as an example.

BACKGROUNDnIn observational vaccine safety studies, diagnosis codes assigned prior to or on the day of vaccination (Day 0) are often excluded from analysis of safety signals since they usually represent pre-existing conditions. The limitations of this approach have been described for autoimmune conditions but not for other chronic conditions. We draw on our experience in a post-licensure quadrivalent human papillomavirus vaccine (HPV4) safety study to examine the effectiveness of exclusion of pre-existing and Day 0 diagnoses of type 2 diabetes mellitus (T2DM) in excluding prevalent T2DM.nnnMETHODSnSubjects included all 117,402 females ages 9-26 years who received HPV4 August 2006-March 2008 in Kaiser Permanente Southern California. We identified potential incident T2DM cases using ICD9 code 250.xx associated with inpatient and emergency room visits during the 60 days following each HPV4 dose, excluding those with this code prior to their first HPV4 dose. Electronic medical records were reviewed to determine the dates of symptom onset, diagnostic labs, vaccine administration and T2DM diagnosis.nnnRESULTSnOf 33 potential incident T2DM cases identified using automated data, 4 (12%) were confirmed to have new onset T2DM after medical record review. Nineteen cases were excluded that did not have T2DM or had T2DM diagnosed before Day 0; nine had an abnormal fasting blood sugar (FBS) ordered on Day 0, prompting subsequent evaluation and diagnosis of T2DM; and one had elevated FBS and glucosuria prior to the first dose of HPV4 but T2DM diagnosed at a visit following vaccination.nnnCONCLUSIONnThese results suggest that among adolescents and young adults, the workup and subsequent diagnosis of pre-existing conditions may result from a visit at which a vaccination is administered. This unmasking phenomenon is not entirely eliminated by exclusion of pre-existing and Day 0 diagnoses. Medical record review should be considered in the evaluation of potential safety signals.

Abstract 647: Use of statin and risk of prostate cancer recurrence among patients who received radical prostatectomy

Proceedings: AACR 103rd Annual Meeting 2012‐‐ Mar 31‐Apr 4, 2012; Chicago, ILnnBackground: HMG-CoA reductase inhibitors (“statins”) have shown broad-spectrum anti-cancer properties in laboratory studies. In epidemiologic studies, statin use has been linked to reduced risk of advanced prostate cancer (PCa). We investigated whether statin use is associated with reduced risk of recurrence in PCa patients who received radical prostatectomy. Methods: All men with incident PCa diagnosed between 2004-2005 who subsequently underwent radical prostatectomy in Kaiser Permanente Southern California (KPSC) health plan were identified using KPSCs accredited cancer registry. Exclusion criteria included 0.2 ng/ml; and (2) clinical disease progression, defined as diagnosis of metastatic disease (identified using ICD-9 codes) or PCa-related death (i.e., PCa as primary cause of death). Information on statin use, demographics, co-morbid conditions, pathoclinical prognostic factors, e.g., stage and Gleason score, and outcomes were obtained from cancer registry and electronic medical records. The effects of statin use prior to and after prostatectomy were both examined using bivariate and multivariable Cox models, adjusting for known prognostic factors. For post-operative statin exposure, a time-dependent Cox model was used. Results: A total of 1192 men were included; 38% had pre-operative and 55% had post-operative statin use. Neither pre- nor post-operative statin use was associated with risk of any outcome, as there was no evidence of a clear dose-response relationship (Table). Conclusion: These data suggest that despite statins protective effect in developing PCa, statin use may not prevent PCa progression.nnnn![Figure][1] nnCitation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research; 2012 Mar 31-Apr 4; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2012;72(8 Suppl):Abstract nr 647. doi:1538-7445.AM2012-647nn [1]: pending:yes

Abstract 3578: The use of preventive services among prostate cancer survivors

Introduction and Objectives: It is well established that most men diagnosed with prostate cancer die from other causes. However, concerns have been expressed that the diagnosis often focuses subsequent care on prostate-related issues rather than overall health. We sought to determine if the use of general preventive services were diminished in a cohort of men following their diagnosis of prostate cancer. Methods: Using the California Men9s Health Study, a prospective cohort study of over 40,000 men ages 45-69 years at baseline in 2002 who were recruited through the Kaiser Permanente Southern California Health Plan, we identified 1636 men who were newly diagnosed with prostate cancer from 2002 through 2008. Preventive services evaluated included screening for colorectal cancer (colonoscopy and/or fecal occult blood tests (FOBT)), tests for diabetes (glucose and hemoglobin A1c) and heart disease (serum cholesterol, high density lipoprotein (HDL) and triglycerides) and vaccinations (influenza and pneumococcal). The use of preventive services was compared in the two years prior to and following prostate cancer diagnosis using matched odds ratios (MOR) and 95% confidence intervals (CI) adjusted for baseline age, Charlson comorbidity index, education, smoking and income. Results: Of the 1636 men diagnosed with prostate cancer, men were equally as likely to receive a pneumococcal vaccination, a hemoglobin A1c or glucose test for diabetes and screening for colorectal cancer before and after prostate cancer diagnosis (Pneumococcal MOR: 1.00, 95% CI: 0.72-1.37; Hemoglobin A1c MOR: 1.01, 95% CI: 0.51-2.01; Glucose MOR: 1.07, 95% CI: 0.71-1.60; Colorectal cancer screening MOR: 0.91, 95% CI: 0.56-1.47). The adjusted odds of receiving an influenza vaccine were 2.6 times greater after prostate cancer diagnosis when compared to before diagnosis (MOR: 2.62, 95% CI: 1.53-4.49). However, men were approximately 70% less likely to receive a cholesterol test (MOR: 0.30, 95% CI: 0.15-0.59) and HDL test (MOR 0.32, 95% CI: 0.16-0.62) following prostate cancer diagnosis. Also, men were approximately 50% less likely to receive a triglyceride test after diagnosis when compared to before prostate cancer diagnosis (MOR 0.51, 95% CI: 0.29-0.89). Conclusions: Results from this study suggest that once diagnosed with prostate cancer, less attention is paid to preventive care for heart disease. As more men with prostate cancer die from causes other than their cancer, screening for and maintenance of preventable diseases of aging is critical and warrants improvement. This analysis also highlights the importance of transitioning care back to primary care physicians after men have undergone treatment for prostate cancer. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research; 2012 Mar 31-Apr 4; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2012;72(8 Suppl):Abstract nr 3578. doi:1538-7445.AM2012-3578