D'Acremont et al identified clinical and laboratory findings associated with imported malaria in returning travelers. This may be helpful when test results are not available or may be unreliable. The authors are from the University of Lausanne and the Swiss Tropical Institute in Basel.
Patient selection:
(1) adolescents or adults (children with malaria tend to present differently, with gastrointestinal findings)
(2) travel to an area with malaria
|
Finding |
Odds Ratio |
Likelihood Ratio for Positive Test |
Likelihood Ratio for Negative Test |
|
splenomegaly |
7.71 |
13.6 |
0.79 |
|
platelet count < 150,000 per µL |
12.4 |
11.0 |
0.43 |
|
hemoglobin < 12 g/dL |
6.92 |
4.6 |
0.88 |
|
poor general health |
4.41 |
3.6 |
0.77 |
|
fever (temperature >= 38°C) |
3.45 |
2.3 |
0.58 |
|
inadequate prophylaxis |
4.65 |
1.6 |
0.35 |
|
sweating |
2.13 |
1.5 |
0.60 |
|
WBC count <= 10,000 per µL |
19.1 |
1.3 |
0.11 |
|
no abdominal pain |
4.05 |
1.2 |
0.58 |
|
eosinophils <= 5% |
4.08 |
1.1 |
0.43 |
from Tables 4 and 5, page 484
where:
• The fever in malaria should be intermittent and not continuous.
• The usefulness of the WBC count, absence of abdominal pain and absence of eosinophilia may be for the exclusion of other infections common in travelers.
• I would think that the decision level for hemoglobin should be different for men and women.
Interpretation:
• The authors multiplied the likelihood ratios to get the overall likelihood for the patient (page 484).
• The authors felt that a posttest probability > 80% could identify patients who might be candidates for presumptive therapy for malaria while definitive tests were pending.
• The clinician can use the posttest probability to help determine test ordering. If the probability is high, then repeating blood smears may be more efficient than hunting for alternative infectious agents.
Specialty: Infectious Diseases
