Paperonium

Cervical cancer · BMJ Glob Health · 2020

Summary: three cervical cancer plans would each prevent 7,700 deaths. Its results, same plans: 10,702.

The summary7,700Its own results section10,70202,0004,0006,0008,00010,00012,000
The summary7,700Its own results section10,70204,0008,00012,000

Cervical cancer deaths prevented by the same three plans, 2020–2040

The paper says

A computer model forecasts cervical cancer deaths among Kenyan women receiving HIV care, 2020–2040. The paper’s summary, the short overview most readers see, says three plans for wider cancer checks and treatment would each prevent about 7,700 of those deaths.

Its own data show

Its results section, on the same three plans and years, says 10,702 deaths prevented. On cancer cases the two agree: about 28,000 prevented. They disagree the same way on a fourth plan, with improved tests: 11,800 deaths prevented in the summary, 15,612 in the results.

“Deaths prevented” is the model’s forecast of deaths without the plans minus its forecast with them. The summary and the results can’t both be right. The results’ figure fits the paper’s forecast of 44,742 deaths without the plans: 10,702 is 23.9% of them, as the paper says. Read against that forecast, the summary understates lives saved by 3,002.

“The summary” means the paper’s abstract: the short summary printed at its top, which is the part most readers see.

Read the paper →Full technical write-up →

The impact of scaling up cervical cancer screening and treatment services among women living with HIV in Kenya: a modelling study · BMJ Glob Health (2020) · doi:10.1136/bmjgh-2019-001886