The standard primipara as a basis for inter‐unit comparisons of maternity care

Publisher Website
Google Scholar
Abstract
Objective To assess the suitability of the standard primipara (a subset of the obstetric population that has relatively low risk or intervention and of adverse outcome) for making inter‐unit comparisons of indicators of the process and outcome of maternity care. Design Inter‐unit comparison of 10 indicators of obstetric intervention and adverse outcome derived from routinely collected computerised data held on the St Mary's Maternity Information System. Setting Fifteen maternity units in the former North West Thames Region. Participants 15,463 primiparae who were delivered in 1992. Main outcome measures Proportion of primiparae within the standard definition; degree to which standard primiparae are associated with lower rates of intervention and adverse outcome, as compared to other primiparae. Results Within the database, 42.6% of all primiparae were found to be standard, with rates varying between units from 25.9% to 57.7%. As expected, the standard primiparous woman is at less risk of intervention or adverse outcome than other primiparae. All but one component variable of the standard definition is a significant risk factor for at least four of the 10 indicators. Statistically significant differences in indicator rates are seen between standard and nonstandard primiparae within units. Within the standard group, significant differences in rates of intervention and adverse outcome are seen between units. Units with relatively high levels of intervention within the higher risk nonstandard group also have relatively high levels of intervention within the standard group. Conclusions Use of the standard primipara, rather than the whole obstetric population, as the basis for inter‐unit comparisons of maternity care will control for the substantial difference in case mix seen in different units, thereby increasing the validity of those comparisons. The technique has the additional benefit of clarifying the relationship between everyday clinical decision making and a unit's performance in comparative indicator reports. The approach must be combined with a separate study of the other groups in the case mix, such as multiparae and high risk primiparae. Additional nonoverlapping groups, homogeneous in terms of risk factors, should be defined and used to extend the basis on which comparisons may be made.