Snapshots of the Literature

Primary and Secondary Syphilis in Baltimore, 1995 [Center for Disease Control, MMWR 45:167,1996]: The CDC reported a 97% increase in the number of primary and secondary syphilis cases reported in 1995 compared to 1993-1994. Demographic information was available for 344 of the 352 cases reported for 1995. Data from the STD clinics showed 27% of patients with syphilis used cocaine compared to 14% without syphilis. Among female patients, prostitution was reported in 35% who reported cocaine use plus syphilis. The seroprevalence of HIV was 8% among 265 STD clinic patients with syphilis compared to 3% of STD clinic patients without syphilis. The CDC concluded that there were at least three possible intersecting epidemics associated with the increased frequency of early syphilis in Baltimore: crack cocaine use, congenital syphilis and HIV infection. Some of this recent increase was possibly ascribed to the reduction in clinicians at the two STD clinics (from 12 to 8) and the number of public health workers conducting contact tracing (decreased from 14 to 8).

By: John G. Bartlett, M.D.

Effects of Incentive and Education Program on Return Rates for PPD Test Reading in Patients with HIV Infection [Chaisson RE, Keruly JC, McAvinue S, Gallant JE, Moore RD. JAIDS 11:455,1996]: This study was a retrospective review of the impact of a food voucher incentive and a patient education program on compliance with PPD reading in HIV-infected patients followed at the Moore Clinic. The standard protocol (control group) was a PPD skin test with the request for a return in 2-3 days for test interpretation. The first intervention consisted of a food voucher for a local fast food restaurant valued at about $4.00 that was given to the patient at the time of the return. The second intervention was a food voucher plus a brief educational message given by the nurse at the time the PPD test was placed. The educational message included an explanation of the importance of the test, the importance of tuberculosis and the interventions that would take place based on the test results. The results are summarized in the table below:

	Number patients	Number returned for reading
Control (no incentive)	272	96 (35%)
Food voucher	229	111 (48%)*
Food voucher + education	96	158 (61%)*
* Difference compared to controls is statistically significant

Multivariant logistic regression analysis showed the factors significantly associated with return for reading were the voucher, the educational message and city residence. Factors that were not significantly associated with return were age, HIV risk factor and race. Of the 365 patients who returned for a reading, 14 (2%) were found to have positive reactions. The conclusion of the authors was that two simple interventions with a combined cost of about $5.00 were associated with a significant increase in the rate of return for PPD readings. This intervention appeared to be cost effective based on the assumption that the cost of treatment of active tuberculosis is approximately $13,000/case.

By: John G. Bartlett, M.D.

Duration of Survival Benefit with AZT Therapy in HIV Infection [Moore RD, Keruly JC and Chaisson RE: Arch Intern Med 156:1073,1996]: This is a study done at the Moore Clinic of 393 HIV infected patients with CD4 cells counts of <500/mm3. The follow-up period was a maximum of three years and a median of two years. There were 235 patients who received AZT for at least 60 days and 158 patients who were non-users (defined as no AZT use or use <60 days). For AZT use during the first 12 months, the relative hazard of death was 0.32, but the survival advantage was no longer statistically significant by year two. The conclusion of the authors was that AZT showed a significant beneficial effect on survival during the first year of therapy, but this benefit is limited in duration to 1-2 years. These data are consistent with multiple other reports showing a time-limited benefit for AZT.

By: John G. Bartlett, M.D.

Chronic Unexplained Diarrhea in Human Immunodeficiency Virus Infection: Determination of the Best Diagnostic Approach [Wilcox CM, et al: Gastro 1996;110:30]: This prospective, 54 month, experience from Grady Memorial in Atlanta is a thorough analysis of chronic (>1 month) diarrhea in 48 patients with HIV infection (90% homosexual, median CD4 18) and repeated negative stool evaluations. All patients underwent upper endoscopy and colonoscopy with extensive histology, including small bowel electron microscopy. Of 21 patients with positive findings, 12 had CMV colitis and 7 had Microsporidia infection. 2 cases of C.difficile and cryptosporidia were identified, one of each combined with CMV colitis. At odds with other experiences, no cases of MAC were found. 12 of 13 findings in the colon could have been made with sigmoidoscopy. A positive diagnosis was more likely to be made in patients with weight loss and lower CD4 counts (only 1 diagnosis made with CD4>50). Survival was significantly worse in these patients (median 8.3 vs. 14.2 months). We agree with both the authors and the accompanying editorial that at least 3 stool cultures and O&P exams and a C. difficile toxin assay remain the cornerstone of evaluation. Beyond that, flexible sigmoidoscopy is of highest yield, especially in patients with colitic symptoms or signs. The enthusiasm for further additional upper bowel biopsy or full colonoscopy should be tempered by the presence of positive predictive factors as described, the severity of symptoms, the presence of comorbid conditions, and the disappointing lack of effective therapy for most pathogens.

By Peter C. Belitsos, M.D.

Biliary Cryptosporidiosis in HIV-Infected People After the Waterborne Outbreak of Cryptosporidiosis in Milwaukee [Vakil NB, et al: NEJM 1996;334:19]: This retrospective study of 82 HIV infected patients acquiring cryptosporidiosis during this largest-ever waterborne epidemic in the US provides the most insight into the natural history of biliary involvement to date. Of this group, 24 patients with biliary symptoms (RUQ pain) were compared to the remainder without. A CD4 count <50/mm3 was a significant risk factor for development of biliary symptoms. Nausea and vomiting were more common (58 vs. 33%), liver-chemistries (especially alkaline phosphatase) became significantly higher, and ultrasound or ERCP frequently demonstrated abnormalities in patients with biliary symptoms. Definitive diagnosis of biliary involvement could be established in 38%, including 4 patients who eventually required cholecystectomy. Increasing age, the presence of nausea and vomiting, and CD4< 50/mm3, were independent predictors of mortality, but the presence or absence of biliary disease was not.

By Peter C. Belitsos, M.D.

Prospective Evaluation of Oropharyngeal Findings in HIV-Infected Patients with Esophageal Ulceration [Wilcox CM, et al: Am J Gastro 1995;90:1938]: The authors identified 124 HIV infected patients with esophageal ulceration out of 343 evaluated by upper endoscopy over a 54 month period and report their oropharyngeal (OP) findings. On presentation with esophageal symptoms, 52% had OP candidiasis. At endoscopy, OP candidiasis remained in 24% and only 14 patients (11%) had an OP ulcer (>3 mm size). OP ulcers were uncommon in patients with CMV (3 of 52) or idiopathic esophageal ulcers (4 of 52) but were common in patients with HSV (6 of 9). Biopsy of OP ulcers were not done, and proved unnecessary, since follow-up after treatment of esophageal disease revealed healing of all OP ulcers. The authors conclude that OP findings are rarely helpful in patients with HIV infection with esophageal ulcer and, when present, are not discriminatory. However, this experience as well as ours, suggests that empiric acyclovir therapy be considered in patients with oral ulceration and odynophagia, reserving endoscopy for failures.

By Peter C. Belitsos, M.D.