From: Establishing seasonal and alert influenza thresholds in Morocco
Concepts | WHO method (5, 20) | MEM (20, 26) |
---|---|---|
Average epidemic curve | Find 3-week moving average of ILI%. Find median peak week for each season. Align the multiple seasons on median peak week. Calculate the average ILI% for each week. Indicates the usual level of influenza activity that occurs during a typical year. | MEM software produces an average curve, lower interval, and higher interval. |
Alert threshold | Calculate the mean and standard deviation (SD) of the average epidemic curve. For each week, the alert threshold is 1.645 SD above the weekly ILI% mean. ILI% > 1.645 SD indicates high ILI activity or outbreaks and may be used to characterize a severe season. | |
Alert curve | A graph consisting of the alert thresholds for each epidemic week. | |
Seasonal threshold (WHO) or pre-epidemic threshold (MEM) | Median weekly ILI% over all weeks (i.e., the average epidemic curve is not used). Indicates the level of influenza activity that signals the start and end of the annual influenza season(s). | For prospective surveillance: upper limit of the 95% one-sided confidence interval of the arithmetic mean of the 30 highest pre-epidemic weekly ILI% values. Parameter value which marks the start of the epidemic period. |
Post-epidemic threshold (MEM) | For prospective surveillance: upper limit of the 95% one-sided confidence interval of the arithmetic mean of the 30 highest post-epidemic weekly ILI% values. | |
Epidemic period start | The third of three consecutive weeks with ILI% above seasonal threshold. Indicates that influenza activity occurs consistently. | For retrospective analysis of individual season data: see “length of epidemic period”. |
Epidemic period end | The third of three consecutive weeks with ILI% below seasonal threshold | For retrospective analysis of individual season data: see “length of epidemic period”. |
Length of epidemic period | Weeks from epidemic start to end. | For retrospective analysis of individual season data: MEM software uses a “maximum accumulated proportions percentage (MAP)” algorithm to split the season into three periods: a pre-epidemic, an epidemic, and a post-epidemic period. |
Epidemic percentage | Proportion of total cases that occurred during the epidemic period | |
Moderate (WHO) or medium (MEM) intensity | Upper 40% limit of 1-sided CI of mean of all peak values. | Upper 40% limit of the one-sided confidence interval of the geometric mean of the 30 highest epidemic weekly ILI% values. |
High intensity | Upper 90% limit of 1-sided CI of mean of all peak values. | Upper 90% limit of the one-sided confidence interval of the geometric mean of the 30 highest epidemic weekly ILI% values. |
Extraordinary (WHO) or very high (MEM) intensity | Upper 97.5% limit of 1-sided CI of mean of all peak values. | Upper 95% limit of the one-sided confidence interval of the geometric mean of the 30 highest epidemic weekly ILI% values. |