Mortality associated with Fournier's gangrene remains unchanged over 25 years.

OBJECTIVES

To report a case series of Fournier's gangrene (FG) from our institution, and to investigate its mortality over the past 25 years.

PATIENTS AND METHODS

Case notes of men presenting to our institution from 2013 to 2016 with FG were reviewed. As well as age, diabetic history, length of stay, length of stay in critical care, and mortality; we calculated (where possible) the Uludag Fournier's Gangrene Severity Index (UFGSI). Published studies and cases series reporting the mortality rates for FG were reviewed from 1993 to 2018. The size of the study, country of origin, average age and gender ratio were collected, alongside mortality.

RESULTS

Two of the 11 patients treated for FG at out institution died within 90 days of admission, a mortality rate of 18%. Predicted mortality was significantly higher. A total of 173 publications were identified from the Medical Literature Analysis and Retrieval System Online (MEDLINE) database published between 1993 and 2018, reporting data from 1975 to 2018. Analysis of heterogeneity, by both time and precision, supported exclusion of four retrospective coded database-driven studies from the analysis. From the remaining studies, mortality ranged from 0% to 42%. Of the 6152 reported cases, there were 1220 deaths, giving an overall mortality rate of 19.8%. There was no evidence of a significant change in the mortality rate for FG over time (P = 0.996).

CONCLUSIONS

In our case series, the mortality rate for FG was 18%, despite a higher predicated mortality (based on UFGSI scores). The treatment of FG remains appropriate resuscitation, aggressive surgical debridement, and critical care management. Perceived high risk of mortality should not deter aggressive management. Mortality due to FG does not appear to have changed over the past 25 years, and is estimated at 19.8%. In studies identifying cases of FG, careful attention should be paid to case definition, particularly when cases are being abstracted retrospectively from large coded databases.