Cases in reported outbreaks, West Bengal

Sum of the 'No. of Cases' printed for the outbreaks counted in 'Disease outbreaks reported' for the week.

Methodology

Source: IDSP's weekly outbreak report, published by the National Centre for Disease Control (NCDC) at ncdc.mohfw.gov.in, one report per reporting week (Monday to Sunday, as printed on the report's cover). Each report lists, state by state, every disease outbreak or alert that States/UTs reported to IDSP's Central Surveillance Unit that week: Unique ID (from 2016), state, district, disease, cases, deaths, date of onset, date of reporting, status and action taken. IHO reads every row of the report's two counted sections: outbreaks reported in the week, and 'Disease outbreaks of previous weeks reported late'. A late-reported outbreak counts in the week whose report lists it, not in the week it began, which is how IDSP's own total (the N printed above the page-2 chart) counts. Rows under 'Follow-up of disease outbreaks of previous weeks' (printed until 2017), and later re-listings of an outbreak already reported in an earlier week (recognised by its earlier Unique ID), update outbreaks already counted and are never counted again. Every state and union territory that existed in the week gets a value, and India gets one: a state with no row in the report gets 0, because the report lists every outbreak reported to IDSP that week, so absence means that none was reported. It does not mean that none occurred. Each week's rows are reconciled before anything is published: the number of rows must equal IDSP's printed N (where printed), every row must carry a distinct Unique ID (before 2016, the roman serials must run without a gap), and every figure in the Cases column must belong to a row. A week that fails is not published. Cases are summed as printed. Where an outbreak's Cases cell gives a breakdown ('92 (village A) and 69 (village B)') the figures are added; where it gives no figure (a dash, 'NA', or a line list in the comments), the outbreak adds nothing. One outbreak listed under several districts adds its districts' cases together.

Caveats

These are outbreaks reported, not disease incidence. The count depends on how many States/UTs sent their weekly report (the cover of each report says how many submitted one, including NIL reports; in some weeks, especially in 2020 and in recent years, fewer than half did), on each state's surveillance capacity and on the threshold its districts apply before calling a cluster an outbreak. A state that reports many outbreaks may simply be one that looks harder. Reporting grew over the years (NCDC counted 554 outbreaks in 2020, 1,862 in 2023 and 3,020 in 2024). The report is published about 5 to 7 weeks after the reporting week ends, so the latest value is never current. Late-reported outbreaks can have begun weeks before the week they count in. The figure is the cumulative number of cases at the time each outbreak was reported, not new cases in the week, and covers only cases linked to a reported outbreak. A single large outbreak (a dengue or cholera outbreak with thousands of cases) can dominate a week. Animal outbreaks (avian influenza in birds) add no human cases.