Clinicial Decision Support in Obstetrics

From Clinfowiki
Jump to: navigation, search

Overview

In obstetrics, clinical decision support needs are unique, as many normal findings are not normal in various stages of pregnancy and vice versa. Since the inpatient obstetrical environment requires constant patient monitoring and can change quickly, the hope is that appropriate clinical decision support has potential for reducing adverse events and improving outcomes.

Outpatient

Early on in the outpatient obstetrics setting, several legacy EHRs such as the STORC (Standard Obstetric Record Charting) system developed by the Madigan Army Medical Center [1], STORC (State Obstetric and Pediatric Research Collaboration) EHR developed by Oregon Health and Science University [2], as well as niche vendors such as eNATAL [3], incorporated CDS elements specific to obstetrics. These included obstetrics format documentation, and gestational age calculators with corresponding alerts, reminders and order sets. Most of the widely used outpatient EHRs from major vendors have obstetrics format documentation and gestational age calculators available, and some have CPOE alerts specific to pregnancy. However, in 2010, commentary from ACOG (the American College of Obstetricians and Gynecologists) criticized the lack of alerts and reminders specific to obstetric management consistent with updated standards of care, and expressed interest in ACOG involvement in EHR certification [4].

Inpatient

Most inpatient obstetrics care involves frequent or constant maternal and fetal monitoring. Patients can be being triaged, admitted for antepartum surveillance, in some stage of labor, or in the postpartum state. The legacy EHRs mentioned previously allowed outpatient data to be incorporated into the inpatient chart, and generated obstetrics formatted documentation from the inpatient data. EHRs which are now widely in use but do not have an obstetrics module at least allow for templates and order sets to be customized to the needs of the obstetrics unit.

In 1995, the company now known as Perigen started with a goal to decrease risk of adverse obstetrics outcomes and reduce malpractice losses [5]. Its EHR is specific to inpatient obstetrics and its CDS features include obstetrics management alerts and documentation. Some alerts relevant to obstetrics include noting blood pressures in the pre-eclamptic range, prolonged second stage of labor, and risks for shoulder dystocia. Documentation is automatically generated with the patient’s inpatient data as well, with reminders for specific pertinent positives and negatives to mention such as pelvic adequacy, no fundal pressure, no traction on infant shoulders, and optional sections for shoulder dystocia and postpartum hemorrhage management, or counseling of patient on unexpected events. These features have mixed reviews from end-users and pose similar challenges as all EHRs with CDS features, such as alert fatigue and difficulty keeping up-to-date with current standards of care or specific hospital policies [6].

Much research has gone into the fetal monitoring component of inpatient obstetrics as well. Fetal heart rate monitoring data can be processed much like telemetry data to detect concerning patterns [7]. Most fetal monitors have built-in alerts for concerning fetal heart rate patterns, although with varying sophistication. In 2009, a company specializing in computerized interpretation of fetal heart rate patterns joined Perigen and began using machine learning using data sets from fetal heart strips of infants with normal and abnormal outcomes to attempt real-time predictive analysis of adverse outcome during labor [8][9]. At the same time, several large studies assessing computer assisted interpretation of fetal telemetry data have been conducted in Europe as well [10][11]. Because of the rarity of permanent infant injury and its ultimately imprecise association with birth data such as APGAR scores and umbilical cord blood gases, it has been difficult to show a statistically significant difference in long-term outcome [12].

References

  1. Nielson PE, et al. Standard obstetric record charting system: evaluation of a new electronic medical record. Obstet Gynecol 2000;96(6):1003-8.
  2. Eden KB, et al. Examining the value of electronic health records on labor and delivery. Am J Obstet Gynecol 2008;199(3):307.
  3. eNATAL – An Internet Based Prenatal Care System. From http://www.enatal.com/index.htm.
  4. McCoy MJ, Diamond AM, Strunk AL. Special requirements of electronic medical record systems in obstetrics and gynecology. Obstet Gynecol 2010;116(1):140-43.
  5. Smith LL, Berry D. Partnering with technology to reduce OB losses. Journal of Healthcare Risk Management 2009;27(4):25-30.
  6. KLAS Research. Perigen PeriCALM Report Card. From https://klasresearch.com/products/reportcard/2145.
  7. Romano M, Bifulco P, Ruffo M, et al. Software for computerised analysis of cardiotocographic traces. Computer Methods and Programs in Biomedicine. 2015; 124: 121-137.
  8. Warrick PA, Hamilton EF, Precup D, et al. Identification of the dynamic relationship between intrapartum uterine pressure and fetal heart rate for normal and hypoxic Fetuses. IEEE Trans Biomed Eng. 2009 Jun;56(6):1587-9.
  9. Warrick PA, Hamilton EF, Kearney RE, et al. A machine learning approach to the detection of fetal hypoxia during labor and delivery. AI Magazine 2012; 33(2): 79-90.
  10. Brocklehurst P. A study of an intelligent system to support decision making in the management of labour using the cardiotocograph— the INFANT study protocol. BMC Pregnancy Childbirth 2016;16:10.
  11. Nunes I, et al. Central fetal monitoring with and without computer analysis: a randomized controlled trial. Obstet Gynecol 2017;129:83–90.
  12. Clark SL, et al. The limits of electronic fetal heart rate monitoring in the prevention of neonatal metabolic acidemia. Am J Obstet Gynecol 2017;216:163.e1-6.


Submitted by Angie Li, MD FACOG