A Comparison of Regulatory Maternity Unit Ratings With Clinical Outcomes and Practice Measures: An Observational Study Using Routinely Collected Data.

OBJECTIVE

To compare inspection-informed ratings of individual maternity units published by the Care Quality Commission (CQC) with clinical outcomes and practice measures.

DESIGN

Observational study using linked national maternity and administrative hospital data.

SETTING

The English NHS.

POPULATION

Women with singleton pregnancies who gave birth at term, April 2018-March 2019.

METHODS

Outcomes and practice measures were compared with ratings using hierarchical models and empirical Bayes estimates adjusted for case-mix and unit characteristics.

MAIN OUTCOME MEASURES

Severe maternal and severe neonatal morbidity. Practice measures included non-spontaneous birth (either caesarean birth before labour or the induction of labour) and intrapartum caesarean birth.

RESULTS

Of 501 719 included women, 39 930 (8.0%) gave birth in 11 units rated 'outstanding', 357 114 (71.2%) in 110 units rated 'good', and 104 675 (20.9%) in 35 units rated 'requires improvement/inadequate'. Severe maternal morbidity did not vary by rating: 1.2% [95% confidence interval 0.87-1.5], 1.3% [1.1-1.4], and 1.0% [0.87-1.1], respectively (p = 0.59), nor did the risk of severe neonatal morbidity: 4.3% [3.3-5.6], 4.0% [3.6-4.5], and 3.4% [2.9-3.9], respectively (p = 0.48). There was no variation across the ratings in the rate of non-spontaneous birth (48.1% [42.2-53.9], 47.9% [46.4-49.4], and 47.9% [45.1-50.8], respectively; p = 0.87) nor intrapartum caesarean (16.8% [14.6-19.3], 16.6% [15.8-17.3], and 15.8% [14.9-16.7], respectively; p = 0.87).

CONCLUSIONS

There was no association between ratings of maternity units published by the national healthcare regulator and clinical outcomes and practice measures derived from routinely collected data. Concerted action is urgently needed to improve the inspection-informed ratings of maternity services.