Paper
Saturday, July 14, 2007
This presentation is part of : Innovations in Women's Health
Nurse-Practitioner Clinical Diagnosis of Vaginal Infections Compared to DNA Probe Analysis
Nancy K. Lowe, RN, CNM, PhD, School of Nursing, Oregon Health & Science University, Portland, OR, USA and Nancy A. Ryan-Wenger, RN, PhD, CPNP, College of Nursing, The Ohio State University, Columbus, OH, USA.
Learning Objective #1: List the components of a systematic diagnostic assessment of vaginal symptoms.
Learning Objective #2: Describe the sources of error in the traditional diagnostic assessment of vaginal symptoms compared to a DNA probe analysis of clinically significant infection.

It is rarely demonstrated empirically that clinician diagnoses are not always accurate even when standard diagnostic protocols are followed. This report focuses on the extent of agreement between clinical diagnoses by advanced practice nurse practitioners (NP) and DNA Probe diagnosis of clinically significant vaginal infection.  Even when protocols are closely followed, diagnoses are not always clear. The APNs in this study were women’s health NP research staff for a project testing the accuracy of women’s self-diagnosis of vaginal symptoms compared to a DNA Probe analysis of vaginal secretions for bacterial vaginosis (BV), trichomoniasis vaginalis (TV), or candida vaginalis (CV). One component of the research protocol was a clinical diagnosis and treatment plan by the research NPs.  The sample was 715 active duty U.S. military women (44% Army, 52.7% Navy, and 3.5% Other Branches), 546 of whom presented for healthcare with vaginal symptoms.  Participants were 38.5% Black, 33.8% White, 18.2% Hispanic or Latino; 41.6% married/cohabiting and 38.1% single; 18 to 54 years (X = 26). The majority was senior (66.4%) or young (25.7%) enlisted personnel. The Clinical Diagnoses (CD) by the NPs included 114 CV (21%), 264 BV/TV (48.6%), 72 mixed BV/CV (13.3%), 48 other (8.8%) and 45 normal (8.3%). The CD had a positive predictive value (PPV) of 80.7% and negative predictive value (NPV) of 74% for BV/TV; a PPV of 67.8% and NPV of 93.2% for CV; a PPV of 53.5% and NPV of 92.7% for Mixed BV/TV/CV infections compared to the DNA standard.  Overall, the NPs correctly diagnosed 83.4% of 320 BV/TV cases, 83.8% of 148 CV cases, and 52.8% of 72 mixed infections.  Further analysis provides insight into the components of the clinical examination related to diagnostic errors. These findings add significant new information to the literature about clinical decision-making in women’s healthcare.