To evaluate the seasonal distribution of retinopathy cases, we calculated the proportion of retinopathy-negative CM cases of the total number of CM cases admitted during each calendar month and compared the proportion in January with each individual month from February to June. Statistical significance of the changes in the odds of retinopathy-negative admissions over months and years was assessed by likelihood ratio tests in logistic regression models including indicator variables for months and years.
A P value less than 0.05 was considered evidence of a significant difference between retinopathy groups. Analysis was performed using S-plus 8.0 (Insightful, Seattle, WA) software.
A total of 2,291 children with a clinical diagnosis of CM were admitted during this time period. Of these children, 1,728 children had determination of retinopathy status. Children who were admitted when an ophthalmologist was not available did not have retinopathy status assessed. Of the 1,728 children with known retinopathy status, 1,056 (61.1%) were malaria retinopathy-positive. The proportion of children with CM who were retinopathy-positive varied throughout the malaria season in parallel with fluctuations of the number of patients presenting with uncomplicated malaria at Queen Elizabeth Central Hospital. When data were combined across calendar years, the proportion of total CM cases that were retinopathy-negative varied significantly between the beginning of the rainy season (January) and its end (June) (P = 0.03) ().
Figure 2. Monthly proportion of children who were retinopathy-negative of the total number of patients admitted with CM in all years combined between 1997 and 2010. The proportion of children admitted with retinopathy-negative CM significantly increased in a linear (more ...)
At the time of year when the incidence of malaria is highest in the community, the proportion of total CM cases that are retinopathy-positive is also at its highest. In contrast, in most years, the annual peak in systemic malaria seen in Malawi is not reflected in a peak in retinopathy-negative CM cases ().
Monthly variation in the number of retinopathy-positive and retinopathy-negative CM patients admitted to Queen Elizabeth Central Hospital from January of 2008 to June of 2010. Histograms were constructed using SAS (SAS Corporation, Cary, NC).
Our findings lend support to the hypothesis that the clinical syndrome of retinopathy-negative CM is unlikely to be associated with acute infection with the malaria parasite. These findings are congruent with data from autopsy studies that show that children dying with retinopathy-negative CM lack cerebral malaria parasite sequestration and have other severe non-malarial illnesses at the time of death.8
Alternatively, the retinopathy-negative CM syndrome could be caused by acute malaria infection with a cofactor (perhaps coinfection) varying seasonally. Additional studies to explore this possibility are warranted.
The half-yearly fluctuation in the proportion of retinopathy-positive CM mirrors the fluctuating incidence pattern seen in uncomplicated malarial illness. This finding reflects the clear increase of malaria transmission intensity that occurs in Malawi in the rainy season. At the time of year when the incidence of malaria is highest in the community, the incidence of retinopathy-positive CM is also at its highest. In contrast, in most years, the annual peak in uncomplicated malaria seen in Malawi is not reflected in a peak in retinopathy-negative CM cases. The proportion of CM patients who were retinopathy-negative increased later in the malaria transmission season.
A limitation of this retrospective study is that patients were recruited for the parent study only during the peak malaria transmission season (January to June each year), and therefore, extrapolation to the remainder of the year is not possible.
We have presented temporal disease pattern evidence supporting the hypothesis that the comas associated with retinopathy-positive and retinopathy-negative CM have different etiologies. The seasonal differences in the proportion of total CM cases that are retinopathy-negative versus retinopathy-positive suggest that the etiologies underlying the retinopathy-negative CM syndrome are distinct from acute severe malaria, the etiology of retinopathy-positive CM.