Use of Obstetric and Gynecologic Hospitalists Is Associated With Decreased Severe Maternal Morbidity in the United States

OBJECTIVES

This study aimed to evaluate the prevalence of obstetric and gynecologic (Ob/Gyn) hospitalists and determine if an association exists between the presence of Ob/Gyn hospitalists and severe maternal morbidity (SMM).

METHODS

This observational study included data from hospitals listed in the USA TODAY 's 2019 article titled, "Deadly deliveries: Childbirth complication rates at maternity hospitals." Telephone and email surveys of staff in these hospitals identified the presence or absence of continuous providers in the hospital 24 hours, 7 days a week (24/7 coverage) and the types of providers who are employed, then compared these responses with the SMM cited by USA TODAY .

RESULTS

Eight hundred ten hospitals were contacted, with participation from 614 labor and delivery units for a response rate of 75.8%. Fifty-seven percent of units were staffed with 24/7 coverage, with 46% of hospitals' coverage primarily provided by an Ob/Gyn hospitalist and 54% primarily by a nonhospitalist OB/Gyn provider. The SMM and presence of 24/7 coverage increased with the level of neonatal care and delivery volume. Of hospitals with 24/7 coverage, those that primarily used Ob/Gyn hospitalists had a lower SMM for all mothers (1.7 versus 2.0, P = 0.014) and for low-income mothers (1.9 versus 2.30, P = 0.007) than those who primarily used nonhospitalist OB/Gyn providers.

CONCLUSIONS

Severe maternal morbidity increases with delivery volume, level of neonatal care, and 24/7 coverage. Of hospitals with 24/7 coverage, units that staff with Ob/Gyn hospitalists have lower levels of SMM than those that use nonhospitalist Ob/Gyn providers.

Comparison of objective, structured assessment of technical skills with a virtual reality hysteroscopy trainer and standard latex hysteroscopy model

OBJECTIVE

To evaluate a virtual reality (VR) hysteroscopy device and a standard latex hysteroscopy model for assessment of residents' surgical skills.

STUDY DESIGN

As part of a 7-station, objective, structured assessment of technical skills (OSATS), 13 obstetrics and gynecology residents performed myoma resection with the Immersion Hysteroscopy VR Trainer (Gaithersburg, Maryland) and dissection of a polyp with a model from Limbs and Things, Inc. (LT) (Bristol, England). OSATS were scored by both blinded (to the residents' levels) and unblinded examiners using a task-specific checklist, global rating scale and pass/fail assessment.

RESULTS

Reliability indices for the VR model were 0.70 (checklist) and 0.98 (global), with an interrater reliability of 0.87 and 0.92, respectively. Correlation of scores between blinded and unblinded examiners was 0.78 for the checklist and 0.88 for the global (p < 0.001). Examination of construct validity revealed improved scores with-increasing postgraduate year. Comparison of VR scores to LT scores revealed VR scores to be significantly higher than LT scores (p < 0.001). Correlations of VR and LT scores were low for both the global (0.478, p = 0.12) and checklist (0.362, p = 0.24).

CONCLUSION

The VR Hysteroscopy Trainer is not superior to standard models for evaluating surgical skills.

Evaluation of p27 in preinvasive and invasive malignancies of the cervix

OBJECTIVE

p27 is a cell cycle inhibitor whose loss is commonly found in epithelial tumors. Low levels have been associated with poor prognosis. Our goal was to determine if p27 expression could be used to screen for dysplasia and if it is a prognostic factor for cervical malignancies.

METHODS

Ten normal cervices, 51 consecutive cone biopsies for preinvasive disease, and 128 consecutive hysterectomies for invasive cervical cancer (1994-1999) were stained for p27 using standard immunocytochemical techniques. All of the cervical cancer patients were managed with radical hysterectomy and lymph node dissection, except for 14 women who underwent adjuvant hysterectomy and lymph node sampling after chemoradiation.

RESULTS

There was no significant difference in p27 staining between normal cervices (all stained 4+) and preinvasive lesions (46/51 stained 4+ and 5/51 stained 3+). For the invasive lesions, 47 women had no residual disease in the hysterectomy specimen, due to prior cone biopsy (41) or radiation (6). All had 4+ p27 staining in the residual cervix. None of these women recurred. Eighty-one women had residual disease in the hysterectomy specimen; 25/81 (31%) had p27 staining of <50%. Loss of p27 was not significantly associated with invasion >50% (32 vs 27%), size >4 cm (20 vs 13%), or use of postoperative radiation (36 vs 20%). Loss of p27 was associated with lymphvascular space invasion (LVSI) (44 vs 20%, P = 0.04). Only 4 women had nodal metastasis; all 4 had p27 staining less than 50%; 6/81 (7%) women with residual disease developed recurrences and died. Of the women who died, 3/6 had p27 staining less than 50%.

CONCLUSION

p27 is not lost in preinvasive cervical lesions and, therefore, cannot be used to screen for dysplasia. In cervical cancers p27 staining was <50% in 31% of cases, and is associated with increased risk of LVSI. Perhaps, because of the excellent overall survival of this group of women with Stage I cervical cancer, loss of p27 staining was not associated with poor prognosis.