BERNALILLO, N.M. – A score of 6 or greater on a sepsis risk scale specifically for pregnant women had an adjusted odds ratio of 109 for ICU admission, with a 95% confidence interval of 18-661, according to its developers at Brown University in Providence, R.I.
Dubbed the Sepsis in Obstetrics Score (SOS), the scale is based on the Rapid Emergency Medicine Score and the SIRS/Sepsis criteria, but is adjusted to reflect the slightly higher heart rates, lower blood pressures, and elevated white counts associated with pregnancy. Scores range from 0-28, with 0 being normal.
Dr. Catherine Albright
The team applied the scoring system retrospectively to 850 pregnant or recently postpartum women who had gotten blood cultures or influenza swabs in the ED; both were used as surrogate markers for septic presentations.
Forty-eight women had scores of at least 6, and eight were admitted to the ICU. There was one ICU admission among the 802 women with scores less than 6.
In addition to the high odds ratio, which was adjusted for age, race, and body mass index, a score of at least 6 had a sensitivity of 88.9%, a specificity of 95.2%, a negative predictive value of 99.9%, and a positive predictive value of 16.7% for ICU admission within 48 hours of presentation.
"In this population in general, there is a low overall rate of serious morbidity and mortality, resulting in our low positive predictive value; it is, however, significantly higher than other disease severity scoring systems studied in this population," said lead investigator Dr. Catherine Albright, an ob.gyn. resident at Brown.
A score of at least 6 also was independently associated with telemetry unit admission, length of hospital stay, fetal tachycardia, and positive blood cultures; the latter were found in 30.8% of women who met the cutoff, but only 8.5% of women who did not. There was no significant difference in the percentage of women with positive flu swabs, about 14% in each group.
"None of the current disease severity scoring systems included pregnant women in their initial study populations. They uniformly overestimate morbidity and mortality in obstetric populations," whereas SOS "can reliably identify patients" who require ICU treatment, Dr. Albright said. The next step is prospective validation, she added.
However, during the question and answer period, an audience member pointed out that the researchers "used your own cohort to develop the cutoff, and then applied that cutoff to the same cohort, and got a high odds ratio; that’s really not surprising. I think your follow-up of applying this to a different cohort [will be] really important."
"Absolutely," Dr. Albright responded.
The score includes temperature, heart rate, blood pressure, respiratory rate, oxygen saturation, white blood cell count, percentage of immature neutrophils, and lactic acid concentration. With adjustment for pregnancy, a heart rate up to 120 bpm, a WBC count of about 6-17,000 cells/mcL, and a systolic blood pressure down to 90 mm Hg are considered to be in the normal range.
Among higher-scoring women, the most common diagnoses at presentation were pyelonephritis and endometritis. Those with scores below six were most commonly diagnosed with influenzalike illness, Dr. Albright reported at the Infectious Diseases Society for Obstetrics and Gynecology annual meeting
There were no significant differences between the two groups in the presence of hypertension, diabetes, HIV, and other comorbidities. The only significant demographic difference was age, with higher-scoring women a mean of 24 years old, vs. 26.3 in the lower-scoring group, probably a clinically irrelevant finding, she said.
Known or suspected ectopic pregnancies and multiple gestations were among the exclusion criteria.
Dr. Albright said she had no relevant financial disclosures.