Table 3.

Summary of methodological, managerial, and evaluative factors in each survey collection model.

Factors		Cohort model	Census model
Methodological factors
	Sample size	Medium to large sample size, predefined	Large, ever-growing sample size with ongoing recruitment
	Representativeness of population	Based on deliveries in birth hospitals	Based on pregnancy at the district level, able to include women from small areas and those who give birth at home or in other settings
	Survey timeliness	Survey at birth requires recall of experiences and outcomes during pregnancy	All surveys relating to the immediate preceding time period
	Bias	Possible sampling bias: enrollment by health professionals after birth may encourage selection of mothers deemed to have had a more positive birth experience	Potential selection bias, although earlier recruitment of mothers reduces the risk of selection based around those deemed to have had a positive birth experience. Selection at first midwife appointment in pregnancy is blind to later experiences and outcomes
	Collection burden	Need for staff training ahead of samples. Enrollment only needed up to a limited period, but is more time-consuming	Ongoing enrollment with less total time spent per health professional. Training only needed for new staff
	Response rate	The initial effective response rate is high, (although lower than that of the census model), with low attrition	The initial effective response rate is the higher of the two models, although drops faster than that in the cohort model
	PROMsa before/after birth	Pelvic floor PROMs are not included as there is no ability to collect prebirth data	Pelvic floor PROMs are included since baseline data at the beginning of the pregnancy are collected
Managerial factors
	Managerial insight	Data provide a snapshot of performance for a certain period of time, enabling lessons to be learned for the following period	Real-time data at different levels of geography enable targeted attention on areas where services need to work better or be better joined up
	Health professional insights	Data provide a snapshot of performance for a certain period of time, enabling lessons to be learned for the following period	Possibility to provide real-time information to different care professionals about the state of delivery of care in their specific area, including highlighting where there are poor experiences or outcomes that professionals could address promptly through their activities
Evaluative factors
	Evaluation models	Enable multidimensional performance assessment	As in the cohort model, and additionally enable inclusion of patient-reported data alongside administrative measures, with contemporaneous reporting periods for both data sets
	Evaluation periods	Data refer to a specific period of collection	Can be used “live” or at any given point in time for evaluating performance
	Analytical approaches	Volume of data can be predetermined according to analytical requirements. Large data sets are possible, enabling advanced statistical models	Continuous collection enables additional analytical approaches (eg, difference in differences) to measure the impact of operational changes

^aPROM: patient-reported outcome.