An evaluation model for syndromic surveillance: assessing the performance of a temporal algorithm.

INTRODUCTION: Syndromic surveillance offers the potential to rapidly detect outbreaks resulting from terrorism. Despite considerable experience with implementing syndromic surveillance, limited evidence exists to describe the performance of syndromic surveillance systems in detecting outbreaks. OBJECTIVES: To describe a model for simulating cases that might result from exposure to inhalational anthrax and then use the model to evaluate the ability of syndromic surveillance to detect an outbreak of inhalational anthrax after an aerosol release. METHODS: Disease progression and health-care use were simulated for persons infected with anthrax. Simulated cases were then superimposed on authentic surveillance data to create test data sets. A temporal outbreak detection algorithm was applied to each test data set, and sensitivity and timeliness of outbreak detection were calculated by using syndromic surveillance. RESULTS: The earliest detection using a temporal algorithm was 2 days after a release. Earlier detection tended to occur when more persons were infected, and performance worsened as the proportion of persons seeking care in the prodromal disease state declined. A shorter median incubation state led to earlier detection, as soon as 1 day after release when the incubation state was < or =5 days. CONCLUSION: Syndromic surveillance of a respiratory syndrome using a temporal detection algorithm tended to detect an anthrax attack within 3-4 days after exposure if >10,000 persons were infected. The performance of surveillance (i.e., timeliness and sensitivity) worsened as the number of persons infected decreased.

[Intestinal paralysis in long-term diabetes mellitus]. / Darmparalyse bei langjährigem Diabetes mellitus.

A 74 year old patient with diabetes mellitus was hospitalized because of nausea, recurrent vomiting and increasing fatigue. Shortly before admittance the patient had diarrhea. He also reported a recent onset of aversion against meat consumption. Clinical investigation revealed a possible right-sided paraumbilical abdominal tumor, normal bowel sounds, a vascular bruit and a normal white blood count with increased band forms. During hospitalisation the general condition of the patient deteriorated rapidly with fever and increasing numbers of immature granulocytes. The patient finally died under the symptoms of a paralytic ileus with hypotonia and hypoglycemia. Autopsy revealed a fist-sized stenosing tumor in the cecum with the histology of a mainly well differentiated, cylindrocellular adenocarcinoma. As immediate cause of death a bilateral paracentral lung embolism with pulmonary edema was found, the latter probably as immediate consequence of preterminal heart failure.

[Urinary specific gravity--comparative measurements using reagent strips and refractometer in 340 morning urine samples]. / Spezifisches Gewicht des Urins--vergleichende Messungen mit Teststreifen und Refraktometer bei 340 Morgenurinproben.

The specific gravity of urine (SG) indicates the number and weight of solute particles in urine; its measurement is helpful in interpreting proteinuria detected by dipstick tests and in monitoring adequate hydration in patients with nephrolithiasis. Four methods for measuring SG or osmolality of urine are currently available (depression of the freezing-point, urometry, refractometry, cation exchange on a reagent strip). Using a recently developed reagent strip, we have measured SG in morning urines of 340 non-selected outpatients and compared the results with SG measurements by refractometry of the same urines. In 86.2% of all urines, a good positive correlation between SG measured by reagent strip and refractometry was noted (r = 0.913, p = 0.0001). In 13.8% of the urines, however, the SG measured by reagent strip deviated by more than +/- 5 from the value obtained by refractometry; in 90% of these urines, glucosuria (reagent strip values too low or too high), proteinuria (values too high), or bacteriuria/leukocyturia (values too low or too high) could be found. In alkaline urine (pH > 7.0), SG values obtained by reagent strip have to be corrected by +5.

Schistosoma haematobium infection in pregnancy.

Due to the economical lack of safe drugs in a remote area of Ghana (Bawku District) to treat Schistosoma haematobium infection during pregnancy, the spontaneous outcome of the pregnancy in women with proved S. haematobium infection was compared with a control group (average hospital delivery). In a survey of 200 pregnant women, we found a prevalence of S. haematobium of 4.5%. From the original collection of 41 infected pregnant women we could follow 23 up to delivery. This group showed a higher number of preterm (less than 37 weeks) deliveries, 34.8% vs. 23.8% in the control group. The birthweights in term deliveries (greater than 37 weeks) were not significantly different (3012 g vs. 3103 g). In the preterm deliveries the birthweight was significantly lower in the infected group (1768 g vs. 2457 g, p less than 0.005).

PIP: Human infection with Schistosome hematobium is common in the Subsahel region of Ghana. Between January 1987 and July 1988, a study was conducted of all pregnant women attending Bawku District Hospital Antenatal-Clinic (ANC), Bawku/Upper East Region (Subsahel-Savanna), Ghana, with complaints and S. hematobium detected in their urine. Pregnant women received iron and folic acid tablets and tetanus-immunization. The partograms of 500 consecutive deliveries in Bawku Hospital (1600 deliveries/year) in the same period were used to collect data for the control group. To estimate the prevalence of S. hematobium infection in pregnancy in Bawku District, 200 pregnant women with a healthy pregnancy were screened. In 9 cases (4.5%) S. hematobium eggs were found in the urine. In 197 cases, the urine test revealed that 51 had proteinuria, 38 had blood in the urine, and 19 women had proteinuria and hematuria. The values were 22-40% with an average of 30%. The hospital delivery analysis of 500 consecutive partograms of the control group showed an average gestation at delivery of 38 weeks, birth weight of 2.917 kg, birth length of 48 cm, and gravidity of 3.6. 41 pregnant women were infected with S. hematobium. One of them had a double-infection of S. hematobium an S. mansoni. Because of 18 drop-outs, the outcome of the pregnancy in 23 women was followed up to delivery. Preterm (37 weeks) deliveries were 34.8% in this group vs. 23.8% in the control group. The births weights in term deliveries (37 weeks) were not significantly different (3012 g vs. 3103 g). In the preterm deliveries, the birth weight was significantly lower in the infected group (1768 g vs. 2457 g, p0.005). The complaints leading to subsequent diagnosis of infection were in 28 cases dysuria, 17 hematuria, 8 waistpain, and 10 lower abdominal pain. In the hospital deliveries, there was one neonatal death in which case the diagnosis of S. hematobium infection was made at the time of the premature delivery.