Volume 8, Number 6—June 2002
Research
Defining and Detecting Malaria Epidemics in the Highlands of Western Kenya
Simon I. Hay*†
, Milka Simba†, Millie Busolo‡, Abdisalan M. Noor†, Helen L. Guyatt*†, Sam A. Ochola‡, and Robert W. Snow*†‡
, Milka Simba†, Millie Busolo‡, Abdisalan M. Noor†, Helen L. Guyatt*†, Sam A. Ochola‡, and Robert W. Snow*†‡Table 2
Comparison of total number of epidemic months detected by the World Health Organization (WHO), Cullen, and cumulative-sum techniques for three study hospitals, western Kenyaa
| Technique | Method | Kilgoris 1998-1999 N=180 (%) | Kisii 1992-2000 N=108 (%) | Tabaka 1986-2000 N=180 (%) |
|---|---|---|---|---|
| WHO | Not transformed | 75 (41.7) | 34 (31.5) | 77 (42.8) |
| Cullen | Not transformed, SCI | 26 (14.4) | 11 (10.2) | 23 (12.8) |
| Not transformed, KCI | 47 (26.1) | 16 (14.8) | 45 (25.0) | |
| Log10 transformed, SCI | 13 (7.2) | 4 (3.7) | 12 (6.7) | |
| Log10 transformed, KCI | 39 (21.7) | 15 (13.9) | 36 (20.0) | |
| C-sum | Not transformed, SCI | 17 (9.4) | 6 (5.6) | 19 (10.6) |
| Not transformed, KCI | 55 (30.6) | 27 (25.0) | 64 (35.6) | |
| Log10 transformed, SCI | 6 (3.3) | 3 (2.8) | 8 (4.4) | |
| Log10 transformed, KCI | 66 (36.7) | 30 (27.8) | 76 (42.2) |
aFigures are number of months defined as epidemic in the monitoring period. Brackets are the percentage of the total months defined as epidemic.
