Surveillance of birth defects: the Multicommunity Sets Technique tested by computer simulation.

Area of Science:

• Medical Surveillance
• Public Health
• Biostatistics

Background:

• Congenital malformations require effective surveillance systems to detect changes in incidence.
• The Multicommunity Sets Technique (MST) and Cumulative Sum (CUSUM) technique are statistical methods used for disease surveillance.

Purpose of the Study:

• To compare the surveillance performance of MST and CUSUM for congenital malformations using computer simulations.
• To evaluate sensitivity, specificity, and detection delay of both techniques under different scenarios of malformation frequency increases.

Main Methods:

• Computer simulations of birth data were used to model congenital malformation occurrences.
• Simulations included increases in malformation frequencies across varying numbers of centers (6/6 and 3/6).
• Performance metrics such as sensitivity, specificity, and detection delay were compared between MST and CUSUM.

Main Results:

• MST demonstrated neither superior sensitivity nor specificity compared to CUSUM.
• MST exhibited a greater delay in signalling increases in malformation frequencies.
• One CUSUM procedure generated more false alarms than MST during prolonged surveillance at baseline rates.

Conclusions:

• CUSUM generally outperformed MST in terms of timely detection of malformation increases.
• The study challenges the presumed suitability of MST for low-incidence malformations, showing CUSUM's advantage.
• These findings have implications for selecting appropriate statistical methods in public health surveillance of congenital anomalies.