Miller, Emily S. MD, MPH; Barber, Emma L. MD; McDonald, Katherine D.; Gossett, Dana R. MD, MSCI
Forceps remain a valuable means to achieve vaginal delivery in cases in which this is unlikely to occur spontaneously.1 As such, forceps represent a tool to curtail the cesarean delivery epidemic.2 Despite this potential, considerable controversy remains regarding the use of forceps and their potential to cause maternal and neonatal injury. The primary maternal complications are severe (third- and fourth-degree) perineal lacerations, which are more frequent in operative vaginal deliveries. In addition, when compared with spontaneous deliveries, forceps-assisted deliveries may be associated with increased risks of various adverse neonatal outcomes such as intracranial hemorrhage.3
Although patient-related factors such as fetal position and station do influence these morbidities, physician characteristics are also likely to be involved. For example, focused obstetrician intervention programs have been developed that demonstrate a reduction in the incidence of some of these injuries.4 One obstetrician characteristic that may be associated with patient outcome is procedural volume. Recent surgical literature has asserted that patient volume is a large determinant of outcome, particularly in highly complex procedures.5–12
Despite the growing recognition of the importance of physician case volume on patient outcomes, data describing this relationship for obstetric procedures are lacking. In particular, patient outcomes after skilled procedures such as forceps delivery may be especially sensitive to experience. Evidence of decreasing experience during residency training and diminishing obstetrician comfort in performing forceps deliveries raise concern about the balance between the risks associated with cesarean and forceps delivery.13 The goal of this analysis was to estimate the effect of attending physician volume of forceps procedures on maternal and neonatal outcomes in a tertiary care teaching hospital. Given recent data in the surgical literature, we hypothesized that obstetricians who perform fewer forceps deliveries annually would have a higher rate of maternal and neonatal adverse outcomes.
MATERIALS AND METHODS
This is a retrospective cohort study of all forceps deliveries performed at Northwestern Memorial Hospital from April 1, 2008, to March 31, 2012. Eligible deliveries were identified through the use of a perinatal database that documents mode of delivery. During this time period, the type of forceps and fetal position began to be routinely recorded as a part of the delivery record. Women were included in this analysis if they were older than 18 years of age, had a singleton gestation, and they either delivered with forceps assistance or if an attempt at forceps was made during their delivery but they ultimately required cesarean delivery.
Once eligible women were identified, their medical records were reviewed. Delivery and neonatal records were abstracted. Potential patient-level confounders for adverse outcomes were identified. These included the following variables previously identified in the literature as associated with severe lacerations: indication for forceps, type of forceps (outlet, low, mid), position of the fetus at application, use of an episiotomy, presence of chorioamnionitis, birth weight, and occurrence of shoulder dystocia.14–17 Charts with missing data were rereviewed by a second investigator; any remaining missing individual data points were excluded from the analysis.
The primary outcome for this study was the occurrence of a severe (third- or fourth-degree) perineal laceration as recorded on the delivery record. At our institution, a third-degree laceration is defined as either a partial or total disruption of the external anal sphincter with intact rectal mucosa. With a two-tailed α of 0.05, this study had 90% power to detect a twofold difference in severe perinatal lacerations between the first and fourth quartiles.
Secondary analyses included adverse neonatal outcomes. Neonatal information was abstracted from the neonatal intensive care unit or normal newborn nursery records and included known neonatal risks associated with forceps delivery, including facial laceration, facial nerve injury, brachial plexus injury, intracranial hemorrhage (subdural, cerebral, intraventricular, and subarachnoid), convulsions, and evidence of central nervous system depression. Other adverse neonatal outcomes were obtained including fracture of the clavicle or humerus, phrenic nerve palsy, or conjunctival or scleral hemorrhage. Given the relative rarity of each of these outcomes, a composite adverse neonatal outcome category was defined as present if any of the aforementioned neonatal outcomes occurred.
Attending physician operative delivery volume was calculated as the number of forceps deliveries performed over the duration of their faculty tenure during the study period. For the purposes of analysis, attending physicians were divided into quartiles by their annual volume of forceps deliveries and this quartile was the primary exposure. In addition, overall attending experience was analyzed as an exposure variable. This was estimated by years in practice at the time of each forceps delivery and was analyzed both as a continuous and a categorical variable. Finally, if a resident was assisting with the delivery, his or her postgraduate year at the time of the delivery was recorded and used as a surrogate for trainee volume. The involvement of a resident was assessed as both a confounder and an effect modifier using a stratified analysis.
Bivariate comparisons between the volume quartiles were performed using the Kruskal-Wallis test, χ2 analysis, or Fisher’s exact test, as statistically appropriate. Factors that were significantly different between comparison groups were considered as potential confounders in the multivariable analysis. This was performed as a multilevel logistic regression model using individual patients at level 1 and attending or resident at level 2 with the “xtlogit” command using Stata 11.1 IC. This study was approved by Northwestern University's institutional review board.
During the study period, there were 118 attending physicians who performed a total of 2,369 forceps-assisted deliveries. Physicians in the lowest quartile of procedure volume performed a median of 1.3 forceps deliveries (interquartile range 1.0–1.8) annually compared with a median of 11.5 forceps deliveries (interquartile range 9.8–17.3) per year in the highest quartile.
Baseline patient-level characteristics of the delivery stratified by attending physician volume quartile are shown in Table 1. Physicians in the highest quartile were significantly more likely to deliver women without a prior vaginal delivery and perform low forceps. In addition, they were less likely to perform an episiotomy at delivery. Indication for forceps assistance was not different across physician procedure volume quartiles, and although fetal position did significantly differ between various groups, there was no significant trend across increasing quartiles using a statistical test for linear trend.
Results of the primary bivariate analysis are depicted in Table 2. Severe perineal lacerations were more frequent among physicians with higher delivery volumes. When analyzing fourth-degree lacerations exclusively, there was no difference in frequency by quartile. Similarly, neonatal outcomes were examined according to attending volume and no difference was noted across the quartiles.
Multilevel multivariable analyses were then performed to estimate whether the association between physician volume quartile and adverse patient outcomes were related to other patient-level characteristics. After adjusting for variables associated with physician quartile in bivariate analysis, prior vaginal delivery, type of forceps, occiput posterior fetal position, and use of median episiotomy remained associated with severe perineal laceration consistent with prior studies. However, once these factors were included, the relationship between physician quartile and severe perineal lacerations disappeared. The adjusted odds ratios are depicted in Table 3.
Physician experience, defined as the number of years since completion of residency, was then examined in multivariable analysis, including the previously identified potential confounders. When analyzed as either a continuous or categorical variable, there was no significant relationship between years of experience and either severe perineal laceration or adverse neonatal outcome (Table 4).
Resident involvement was then examined in a similar manner. Postgraduate year was not significantly associated with the indication for forceps, station at application of forceps, fetal position, or the use of an episiotomy (Table 5). After controlling for these potential patient-level confounders, there was no association between severe perineal laceration and resident level of training (Table 6). However, when compared with first-year residents, third- and fourth-year residents had a lower odds of experiencing a composite adverse neonatal outcome.
Finally, an exploratory analysis was performed to examine whether a threshold physician volume could be identified below which severe perineal lacerations became more frequent. Annual forceps volume thresholds between two and 10 were analyzed and no threshold was identified below which rates of severe lacerations or composite adverse neonatal outcomes were increased.
Operative vaginal delivery rates across the United States have declined.18 This decrease has already trickled down to affect trainees; many graduating obstetrics residents do not feel comfortable independently performing a forceps-assisted delivery.19 According to the Accreditation Council for Graduate Medical Education, the median number of forceps performed before graduation was six.20 With dwindling procedure numbers, the question arises of how this diminishing volume will affect patient outcomes.
Our study suggests that physician forceps volume does not affect rates of severe perineal laceration. Although our statistical power was low to detect differences in this outcome, we did not observe any relationship between the risk of composite neonatal injury and obstetrician volume. A threshold minimum of forceps deliveries that led to an increase in adverse outcomes could not be identified. The lack of an association between obstetrician volume and adverse outcomes persisted after controlling for patient variables associated with adverse sequelae. The lack of an association persisted even when analyzed by health care provider experience in multivariable analysis.
To identify whether volume during earlier experiences influences patient outcomes, we performed an exploratory analysis using resident physician year of training. After controlling for confounders, the year of residency, a surrogate for physician volume, was not associated with rates of severe perineal lacerations. However, interestingly, the odds of composite adverse patient outcomes were lower among third- or fourth-year residents compared with interns. Thus, our data suggest that for forceps deliveries, adverse outcomes are largely attributable to previously identified patient characteristics such as fetal station and position. However, these data also indicate that there may be some association between composite adverse neonatal outcomes and early training experiences.
This study is not without methodologic limitations. First, these data are derived from a single institution with a unique delivery profile. There were 118 physicians who used forceps assistance during the study time period; only eight of whom delivered less than one patient per year by forceps. The absence of a relationship between physician volume or years of experience and adverse patient outcome may be related to the unique practice patterns of our obstetricians compared with the majority of American obstetricians. A threshold analysis was used in an attempt to identify if a limit below which adverse outcomes increased without any identified change. Similarly, the majority of physicians performing forceps deliveries trained in an institution or an era where forceps delivery was more frequent; thus, their self-perceived competency may be high. Patient outcomes may be related to the overall experience of a particular attending and thus a lifetime threshold, rather than an annual volume threshold, may be the physician–volume measure most associated with patient outcomes. This may be supported by the decreased odds of adverse neonatal outcomes associated with increasing years of residency training.
Finally, although we attempted to control for confounders using identifiable patient-level characteristics, there are clearly variations of difficulty even within these categories; it may be that more experienced health care providers attempt more difficult operative vaginal deliveries. This omitted variable bias may mitigate the direct relationship between forceps volume and severe perineal laceration observed in bivariable analysis, but its influence is likely smaller than that of the included confounders. Thus, it is unlikely to affect the results of the multilevel multivariable regression.
From a policy and patient safety perspective, it is important to identify sources of variation in outcomes to optimize quality of care. The alternative to an operative vaginal delivery is, of course, cesarean delivery, which is not without its own risks both in the index pregnancy as well as in all future pregnancies. Because of these morbidities, reducing the prevalence of cesarean delivery remains a national goal.21 Thus, forceps delivery must remain in our obstetric armamentarium as a tool to reduce the use of cesarean delivery and identifying practices to optimize patient safety are critical. Although attending obstetrician volume or experience did not explain variation in patient outcomes in this study, increased experiences during residency did demonstrate an association with improved neonatal outcomes. This finding underscores the importance of including closely supervised forceps experience during residency training to optimize patient outcomes in the future.
1. Operative vaginal delivery. ACOG Technical Bulletin No. 196—August 1994 (replaces No. 152, February 1991). American College of Obstetricians and Gynecologists. Int J Gynaecol Obstet 1994;47:179–85.
2. Freeman RK, Cohen AW, Depp R, Frigoletto FD, Hankins GDV, Lieberman E, et al.. Evaluation for cesarean delivery. Washington (DC): ACOG Task Force on Cesarean Delivery Rates; 2000.
3. Towner D, Castro MA, Eby-Wilkens E, Gilbert WM. Effect of mode of delivery in nulliparous women on neonatal intracranial injury. N Engl J Med 1999;341:1709–14.
4. Hirsch E, Haney EI, Gordon TE, Silver RK. Reducing high-order perineal laceration during operative vaginal delivery. Am J Obstet Gynecol 2008;198:668.e1–5.
5. Begg CB, Riedel ER, Bach PB, Kattan MW, Schrag D, Warren JL, et al.. Variations in morbidity after radical prostatectomy. N Engl J Med 2002;346:1138–44.
6. Birkmeyer JD, Siewers AE, Finlayson EV, Stukel TA, Lucas FL, Batista I, et al.. Hospital volume and surgical mortality in the United States. N Engl J Med 2002;346:1128–37.
7. Birkmeyer JD, Stukel TA, Siewers AE, Goodney PP, Wennberg DE, Lucas FL. Surgeon volume and operative mortality in the United States. N Engl J Med 2003;349:2117–27.
8. Schrag D, Panageas KS, Riedel E, Hsieh L, Bach PB, Guillem JG, et al.. Surgeon volume compared to hospital volume as a predictor of outcome following primary colon cancer resection. J Surg Oncol 2003;83:68–78.
9. Schrag D, Earle C, Xu F, Panageas KS, Yabroff KR, Bristow RE, et al.. Associations between hospital and surgeon procedure volumes and patient outcomes after ovarian cancer resection. J Natl Cancer Inst 2006;98:163–71.
10. Brennan MF, Radzyner M, Rubin DM. Outcome—more than just operative mortality. J Surg Oncol 2009;99:470–7.
11. Gruen RL, Pitt V, Green S, Parkhill A, Campbell D, Jolley D. The effect of provider case volume on cancer mortality: systematic review and meta-analysis. CA Cancer J Clin 2009;59:192–211.
12. Juillard C, Lashoher A, Sewell CA, Uddin S, Griffith JG, Chang DC. A national analysis of the relationship between hospital volume, academic center status, and surgical outcomes for abdominal hysterectomy done for leiomyoma. J Am Coll Surg 2009;208:599–606.
13. Powell J, Gilo N, Foote M, Gil K, Lavin JP. Vacuum and forceps training in residency: experience and self-reported competency. J Perinatol 2007;27:343–6.
14. deLeeuw JW, de Wit C, Kuijken JP, Bruinse HW. Mediolateral episiotomy reduces the risk for anal sphincter injury during operative vaginal delivery. BJOG 2008;115:104–8.
15. Bofill JA, Rust OA, Devidas M, Martin RW, Morrison JC, Martin JN. Prognostic factors for moderate and severe maternal genital tract laceration with operative vaginal delivery. Am J Obstet Gynecol 1996;174:353.
16. FitzGerald MP, Weber AM, Howden N, Cundiff GW, Brown MB; Pelvic Floor Disorders Network. Risk factors for anal sphincter tear during vaginal delivery. Obstet Gynecol 2007;109:29–34.
17. Ecker JL, Tan WM, Bansal RK, Bishop JT, Kilpatrick SJ. Is there a benefit to episiotomy at operative vaginal delivery? Observations over ten years in a stable population. Am J Obstet Gynecol 1997;176:411–4.
18. Martin JA, Hamilton BE, Ventura SJ, Osterman MJ, Kirmeyer S, Mathews TJ, et al.. Births: final data for 2009. Natl Vital Stat Rep 2011;60:1–70.
19. Bofill JA. Operative obstetrics: a lost art? Obstet Gynecol Surv 2000;55:405–6.
20. Accreditation Council for Graduate Medical Education. Available at: http://www.acgme.org/
. Retrieved September 21, 2013.
© 2014 by The American College of Obstetricians and Gynecologists.