Transcript Document
Enough to Make You CRiNGe: Variation in Adherence to the Treatment Guidelines for Neisseria gonorrhoeae, California, 2009-2011 Richard Lechtenberg, MPH University of California, Berkeley California Dept. of Public Health Gonorrhea: Nothing to Clap About • 2nd most common reported infectious disease • Risk factor for… – Pelvic inflammatory disease – Ectopic pregnancy – Infertility • Facilitates transmission and acquisition of HIV • Largest racial health disparities • Antimicrobial resistance 2 Study Question • How does adherence to the CDC treatment guidelines for gonorrhea vary by clinical practice setting? 3 The California Gonorrhea Surveillance System Provider diagnosis Positive lab test Case report Lab report Local health jurisdiction Electronic submission to the state Small LHJs Random sample of cases drawn Pre-populated interview record sent via secure e-mail Patient and provider interviewed by phone Data entered at the state 4 Pharynx Cervix, Urethra, & Rectum Adherent Treatment 8/4/2006 – 12/16/2010 12/17/2010 – 8/9/2012 Ceftriaxone 125 mg IM OR Cefixime 400 mg Ceftriaxone 250 mg IM OR, IF NOT AN OPTION Cefixime 400 mg PLUS Azithromycin 1g OR Doxycycline 100 mg BIDx7 Ceftriaxone 125 mg IM Ceftriaxone 250 mg IM PLUS Azithromycin 1g OR Doxycycline 100 mg BIDx7 lag 12/16/10 2006 guidelines in effect 2010 guidelines in effect Analysis Inclusion and exclusion criteria related to specific diagnoses Bivariate associations tested using weighted χ2 tests Directed acyclic graphs (DAG) used to identify confounders Independent cumulative incidence ratios (CIR) estimated using weighted generalized linear models Adjusted Wald tests used to test the significance of sets of coefficients (α=0.10) 6 Software • Dataset compiled in SAS 9.2 • Analyses in R 2.13.1 • DAG in daggity.net 7 Percent receiving a non-rec. treatment by clinical setting Military/VA Correctional facility ER/Urgent Care Other Hospital Private physician/HMO Family planning facility Comm./Pbl Hlth clinic HIV clinic STD clinic 0% 5% 10% 15% 20% 25% 30% 35% 40% 9 Causal Diagram & Modeling Process p=0.48 11 12 13 14 15 Summary and Recommendations • >2-fold increase in risk of non-rec. tx in nearly all settings compared to STD clinics; esp. high at... – Military/VA facilities – Correctional facilities • But intervention may be most fruitful at… – family planning facilities – private physicians/HMOs 16 Summary and Recommendations (contd.) • Lower risk of receiving a non-rec. tx among… –MSM –Blacks and Hispanics • Directions for future research –Identification of specific barriers to the provision of guideline-concordant treatment 17 Limitations Strengths Robust results •Robust to sensitivity analyses Information Bias Sample size •Misclassification of the outcome •Large Selection Bias Data •Response Bias •Reporting Bias •variety of clinical settings •Large geographic region 18 Acknowledgements • Colleagues at CDPH – Michael Samuel, DrPH* – Heidi Bauer, MD, MS, MPH* – Joan Chow, DrPH, MPH – Ina Park, MD, MS – Nicole Olson, MPH* – Scott Baker, MPH – Jessica Frasure-Williams, MPH – Mary Fredrickson – Carol Kong, MPH* • My professors at UC Berkeley – – – – Kyle Bernstein, PhD, ScM* Maureen Lahiff, PhD* Barbara Abrams, DrPH Jack Colford, MD, PhD Contact Info Richard Lechtenberg, MPH [email protected] *Co-authors 19