Evaluating the performance of the M6 model without initial serum progesterone (M6NP) and the NICE guidelines in a Danish pregnancy of unknown location (PUL) population.

OBJECTIVE

Prediction models based on serum human chorionic gonadotrophin (s-hCG) levels are used to risk stratify women with pregnancy of unknown location (PUL) into low and high risk for ectopic pregnancy (EP). The objective of this study was to evaluate the performance of the M6NP model and the NICE guidelines in a population of Danish women with PUL. The secondary objective was to assess initial ultrasound findings as a risk stratification tool.

METHODS

This two-center retrospective cohort study included 773 women with PUL. Clinical presentation, final diagnosis, s-hCG measurements and ultrasound characteristics at all visits were recorded. The M6NP model and the NICE guidelines were retrospectively applied both within and outside their recommended timeframes.

RESULTS

The M6NP model demonstrated an overall sensitivity of 97 % and specificity of 46 % while the NICE guidelines demonstrated an overall sensitivity of 91 % and a specificity of 66 %. The M6NP model misclassified two cases of EP compared with six misclassified by the NICE guidelines. However, the higher sensitivity of the M6NP model came at the cost of increased surveillance, with 54 % of women classified as high risk for EP, compared to 39 % with the NICE guidelines. An initial ultrasound indicating a probable EP was associated with a 24 % risk of EP, whereas a probable IUP was associated with an EP risk of 2.7 %.

CONCLUSION

Both the M6NP model and NICE guidelines demonstrated reliable performance in stratifying the risk of EP and guiding appropriate follow-up protocols for women with PUL. Initial ultrasound findings are a valuable tool for classifying PUL cases into low, intermediate, and high-risk groups. Combining initial ultrasound with a validated risk stratification model may improve the overall management of PUL.