The M6 risk prediction model and two-step strategy to characterize
pregnancies of unknown location: a multicentre external validation study
Abstract
Objective. To externally validate the M6 risk model and the two-step
triage strategy (2ST) to triage pregnancies of unknown location (PUL),
and compare performance with the M4 model and beta human chorionic
gonadotropin (BhCG) ratio cut-offs. Design. Model validation study.
Setting. Eight UK hospitals with early pregnancy assessment units.
Population. Women presenting with a PUL and BhCG >25 IU/L.
Methods. Women were managed using the 2ST protocol: step 1 classifies
PUL as low risk of ectopic pregnancy (EP) if presenting progesterone ≤2
nmol/L, M6 is used as step 2 in the remaining cases. We validated 2ST
and M6 alone (with and without progesterone as a predictor: M6P and
M6NP). M6 and M4 require the BhCG ratio over two days. Based on these
models, we classified PUL as high risk for EP when the risk was ≥5%. We
meta-analysed centre-specific results. Main outcome measures.
Discrimination, calibration and clinical utility (decision curve
analysis) for predicting EP. Results. Of 2899 eligible women, the main
analysis excluded 297 (10%) women that were lost to follow-up. 16%
(95% confidence interval 12-20) of women had presenting progesterone ≤2
nmol/L. The area under the ROC curve for EP was 0.88 (0.86-0.90) for 2ST
and 0.89 (0.86-0.91) for M6P. Sensitivity for EP was 94% (89%-97%)
for 2ST and 96% (91%-98%) for M6P. Both approaches had good overall
calibration, with modest variability between centres. M4 and BhCG ratio
cut-offs had inferior performance and lower clinical utility.
Conclusions. The 2ST and M6P alone are the best approaches to triage
PUL.