Oliphant, Sallie S. MD; Jones, Keisha A. MD; Wang, Li MS; Bunker, Clareann H. PhD; Lowder, Jerry L. MD, MSc
From the Division of Urogynecology, Department of Obstetrics, Gynecology, and Reproductive Sciences, Magee-Womens Hospital, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania; Office of Clinical Research, University of Pittsburgh Clinical and Translational Science Institute, Pittsburgh, Pennsylvania.
Supported by grant number UL1 RR024153 from the National Center for Research Resources (NCRR), a component of the National Institutes of Health (NIH), and NIH Roadmap for Medical Research. Its contents are solely the responsibility of the authors and do not necessarily represent the official view of NCRR or NIH. Information on NCRR is available at http://www.ncrr.nih.gov/.
Presented as a poster at the 30th annual American Urogynecologic Society Scientific Meeting, September 24–26, 2009, Hollywood, Florida.
Corresponding author: Sallie S. Oliphant, MD, Division of Urogynecology, University of Pittsburgh School of Medicine, Magee-Womens Hospital, 300 Halket Street, Pittsburgh, PA 15213; e-mail: firstname.lastname@example.org.
Financial Disclosure The authors did not report any potential conflicts of interest.
Women in the United States commonly undergo obstetric and gynecologic procedures, such as vaginal delivery, cesarean delivery, and hysterectomy, during inpatient hospital stays. Little data exist regarding trends in these procedures. Rutkow et al1 published data in 1986 examining obstetric and gynecologic surgery trends from 1979 to 1984. According to this study, obstetric and gynecologic procedures comprised 5 of the top 10 most common surgical procedures in the United States.1 Other studies have examined trends in specific gynecologic and obstetric procedure trends, but no study has looked at recent trends in overall procedures.2–4
The purpose of this study was to estimate trends over time in all major inpatient obstetric and gynecologic procedures using data from the National Hospital Discharge Survey from 1979 to 2006. A second aim was to estimate principal inpatient obstetric and gynecology surgical procedures performed across the lifespan of U.S. women by decade of life year.
MATERIALS AND METHODS
Data were abstracted from the National Hospital Discharge Survey, a federal dataset using a two-stage (1965–1987) and a three-stage (1988–2006) probability sampling of hospitals, followed by systematic random sampling of inpatient hospital discharges within these hospitals. Staging was designed to obtain representative distribution by geographical location, bed size, and type of ownership. The 1965 group of hospitals was followed-up through 1987 with periodic addition of new hospitals to replace those that closed or became ineligible. Using a similar approach, a new set of hospitals was selected in 1988.5 Medical records from 466 nonfederal short-stay hospitals (8% of all U.S. hospitals) were selected and approximately 270,000 discharges were collected per year from January 1979 to December 2006, the current publicly available time period. The survey recorded up to seven discharge diagnosis codes and four procedure codes using the International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM) coding system. Other de-identified information collected included patient gender, age, race, marital status, length of hospital stay, hospital size (number of beds), hospital ownership, and insurance type or expected source of payment. Quality-control programs have estimated the error rate for the National Hospital Discharge Survey at 4.3% for medical coding and data entry, and 1.4% for demographic coding and data entry.6,7
After obtaining University of Pittsburgh Institutional Review Board approval (exempt status), women who underwent obstetric or gynecologic surgical procedures from 1979 to 2006 were identified using ICD-9-CM procedure codes. Table 1 lists the broad coding categories selected for study. All women undergoing one or more of these procedures were included in the analysis. Weights provided by the National Hospital Discharge Survey for each patient discharge were used to tabulate surgical procedure numbers and demographics to allow inflation to national estimates. For patients undergoing multiple procedures, these procedures were counted separately in their respective categories, but for multiple procedures they were counted under the same broader category, and for demographic data an individual patient was tallied only once. Age-adjusted rates of procedures per 1,000 women were calculated by the direct method of rate adjustment using the 1990 projected U.S. Census population data for each year of age. Use of age-adjusted rates based on a standard population permitted examination of trends independent of changes in population age structure over time. Data from the 1990 U.S. Census were chosen to represent a midpoint in the study time period. Standard error of the age-adjusted rate was calculated by the method provided by the National Hospital Discharge Survey.5 When the estimated number of cases per year was based on fewer than 60 records in the database, the estimate was considered unreliable.6 Statistical analysis was performed using SPSS 17.0 software.
From 1979 to 2006, approximately 137,128,000 obstetric and gynecologic inpatient procedures were performed in the United States. Obstetric and gynecologic procedures comprised 26.5% of inpatient surgical procedures in adult women during the 28-year study time period. Table 2 lists the top 10 most frequently occurring inpatient procedures based on three-digit ICD-9-CM procedural code classification for years 1979 to 2006. In 1979, 7 of these top 10 procedures were obstetric or gynecologic in nature. In 2006, 6 of the top 10 were obstetric and gynecologic in nature.
Of the more than 137 million identified procedures identified, 64% were solely obstetric and 29% were solely gynecologic, with 7% of women undergoing both an obstetric and a gynecologic procedure during the same inpatient admission. The mean age of women undergoing an obstetric procedure was 27±6 years, and 39±14 years for a gynecologic procedure. Overall length of hospital stay for obstetric procedures was 2.9±4.8 days, and for gynecologic procedures it was 4.0±4.8 days. Table 3 lists age and length of stay for women undergoing obstetric and gynecologic procedures by study year. Length of stay for cesarean delivery decreased from 6.5±3.5 days in 1979 to 3.6±3.2 days in 2006. Length of stay for hysterectomy also decreased from 8.5±4.3 days in 1979 to 2.8±3.5 days in 2006. The number of women undergoing inpatient obstetric and gynecologic procedures has decreased over time from 5,351,000 in 1979 to 4,949,000 in 2006. For gynecologic procedures, the number of procedures has decreased by 46%, from 2,852,000 in 1979 to 1,309,000 in 2006. For obstetric procedures, the number has increased 1.5-fold, from 2,755,000 in 1979 to 4,014,000 in 2006.
Figure 1 depicts trends over time of obstetric procedures using age-adjusted rates. Age-adjusted rates for spontaneous vaginal delivery have increased slightly, from 25.2 per 1,000 women (±standard error [SE] 1.58) in 1979 to 26.0 per 1,000 women (±SE 1.63) in 2006. Age-adjusted rates for operative vaginal delivery have declined more than twofold, from 6.1 per 1,000 (±SE 0.47) in 1979 to 2.9 per 1,000 (±SE 0.20) in 2006. Episiotomy age-adjusted rates decreased by more than 75% during the study period, from 20.2 per 1,000 (±SE 1.34) in 1979 to just 4.6 per 1,000 (±SE 0.31) in 2006. Cesarean delivery age-adjusted rates doubled over the time period, from 6.4 per 1,000 (±SE 0.49) in 1979 to 13.5 per 1,000 (±SE 0.86) in 2006.
Overall age-adjusted rates for gynecologic procedures decreased during the study period, with the exception of procedures for stress urinary incontinence, which increased (Fig. 2). Age-adjusted rates of abdominal and laparoscopic hysterectomies decreased from 6.3 per 1,000 women (±SE 0.49) in 1979 to 3.1 per 1,000 (±SE 0.21) in 2006. Age-adjusted rates for vaginal hysterectomy declined from 2.2 per 1,000 (±SE 0.19) in 1979 to 1.5 per 1,000 (±SE 0.11) in 2006. Age-adjusted rates for oophorectomy decreased from 5.7 per 1,000 women (±SE 0.45) in 1979 to 3.3 per 1,000 (±SE 0.22) in 2006. Age-adjusted rates for sterilization procedures decreased by half, from 7.6 per 1,000 (±SE 0.58) in 1979 to 3.8 per 1,000 (±SE 0.26) in 2006. Age-adjusted rates of prolapse procedures decreased from 2.7 per 1,000 (±SE 0.24) in 1979 to 1.3 per 1,000 (±SE 0.10) in 2006, whereas rates of incontinence procedures increased from 0.8 per 1,000 (±SE 0.08) in 1979 to 0.9 per 1,000 (±SE 0.08) in 2006.
Figure 3 depicts procedures performed in 2006 in women divided into age groups by decade. The most common procedure performed in women age younger than 40 was spontaneous vaginal delivery. For women age 40 to 69 years, the most common procedure was hysterectomy. In the age 70 years and older group, the most commonly performed procedure was a repair for pelvic organ prolapse.
Inpatient gynecologic surgical procedures have decreased dramatically from 1979 to 2006, both with regard to total number of procedures and percentage of overall inpatient surgical procedures for adult women, whereas inpatient obstetric surgical procedures have increased. Overall, obstetric and gynecologic procedures still comprise a large proportion of overall surgical volume for adult women in the United States. As expected, the type of surgery a woman is likely to undergo is strongly correlated with age and reproductive stage of life.
Surgeries traditionally performed by obstetrician-gynecologists comprised more than 25% of inpatient surgical procedures for U.S. women over the course of the 27-year study period. In addition, out of all inpatient surgical procedures performed in women, obstetric- and gynecologic-related procedures comprised the majority of the top 10 procedures across the study period. This data set does not record surgeon subspecialty; therefore, some of these traditionally obstetric and gynecologic procedures could have been performed by other subspecialists (ie, family medicine, general surgery). The decrease in inpatient gynecologic procedure trends over the study period likely reflects changing practice patterns, including increased use of minimally invasive outpatient surgeries and advances in medical treatments offered in place of surgery.8 The overall rate of hysterectomy has decreased dramatically. In a recent study examining trends in hysterectomy with oophorectomy, the rate of hysterectomy alone decreased from 1979 to 2004. Although the rate of hysterectomy with oophorectomy decreased in women younger than 50 years, the rate of hysterectomy with oophorectomy increased in women age 50 years and older.9,10 The rate of procedures for treatment of stress urinary incontinence increased over the study period and is likely attributable to the introduction of effective, minimally invasive procedures such as the synthetic midurethral sling.3 This trend will likely continue as the proportion of the population affected by stress urinary incontinence is projected to increase dramatically over the next 30 years.11 Although the overall rate of surgical procedures for prolapse decreased, surgical repair of prolapse was the most common inpatient procedure performed in women older than 70 years. Jones et al,12 in a recent study of surgical treatment of prolapse procedures, noted decreasing rates in women younger than 52 years, whereas in women age 52 years or older the rates remained stable, likely reflecting the true rate of prolapse procedures in the United States.
Although operative vaginal delivery procedure rates have dramatically decreased, rates of cesarean delivery increased considerably. Martin et al13 showed that the rate of cesarean delivery has increased steadily since 1996, reaching a record 31.1% in 2006. The dramatic increase in the cesarean delivery rate likely reflects changing obstetric practice patterns, including an increase in the rate of primary cesarean delivery, a decline in the rate of vaginal birth after cesarean delivery, and a decline in the use of operative vaginal delivery. Studies have shown a dramatic decrease in use of episiotomy and operative vaginal deliveries, which appears to inversely correlate with increasing cesarean delivery rates, possibly representing physicians' route of delivery choice for more difficult deliveries.13,14 The decrease in rates of episiotomy may also reflect practice changes attributable to recognition of increased rates of posterior perineal trauma with routine episiotomy use.15
The U.S. adult female population is estimated to increase by nearly 50% by 2050.16 Current projections estimate that the ob-gyn physician workforce will increase by only 15% by 2020.17 In 1996, there were 38,424 U.S. physicians registered by specialty type as ob-gyns, increasing to only 42,333 in 2006. Despite an overall increase in the number or ob-gyn physicians over the study time period, the number of ob-gyns relative to the total population has not changed significantly since 1994. According to American Medical Association data, the number of ob-gyns per 10,000 U.S. citizens was 14.1 in 1994, remaining 14.1 per 10,000 in 2006 (range 14.1–14.7).17 Although overall inpatient surgical procedures, specifically gynecologic procedures, have decreased over time, obstetric and certain gynecologic procedures have increased. It is crucial to maintain an adequate ob-gyn subspecialist workforce to provide essential surgical services to our female patients.
These data are based on cross-sectional sampling and are dependent on both accuracy of coding and generalizability of the sampling. The National Hospital Discharge Survey data have been shown to have reasonable coding accuracy.6 We chose U.S. Census data from 1990, a time point midway through the study, to calculate the age-adjusted rates. Census data from 1990 may not accurately reflect year-to-year population change. However, data from a single year were needed to base and compare age-adjusted rates across the entire study time span. We recognize that the 1988 National Hospital Discharge Survey sampling redesign may have influenced the overall trends; however, examination of frequencies for each year by procedure (data not shown) revealed no consistent change in rates from 1987 to 1988.
Additionally, the National Hospital Discharge Survey underestimates the number of surgical procedures in U.S. women because it does not collect data from same-day surgeries or ambulatory surgical centers. The lack of data on same-day surgeries will continue to affect the accuracy of estimates of total U.S. surgical procedures as an increasing number of surgical procedures are being performed on an outpatient basis. The National Survey of Ambulatory Surgery, a federal database similar to the National Hospital Discharge Survey, collected data on ambulatory surgical procedures from 1994 to 1996 and was then discontinued. The National Survey of Ambulatory Surgery was reinstituted during 2006 for a 1-year period. Continued ambulatory surgical data collection, in addition to inpatient data collection, is needed to accurately assess future surgical trends.
The National Hospital Discharge Survey provides a large sample size, well-defined inclusion criteria, and standardized use of diagnosis and procedure codes. Although inpatient obstetric and gynecologic procedures rates decreased from 1979 to 2006, obstetric and gynecologic procedures still represent a large proportion of surgical volume for women in the United States.
1. Rutkow IM. Obstetric and gynecologic operations in the United States, 1979 to 1984. Obstet Gynecol 1986;67:755–9.
2. Kozak LJ, Weeks JD. U.S. trends in obstetric procedures, 1990–2000. Birth 2002;29:157–61.
3. Oliphant SS, Wang L, Bunker CH, Lowder JL. Trends in stress urinary incontinence inpatient procedures in the United States, 1979–2004. Am J Obstet Gynecol 2009;200:521.e1–6.
4. Frankman EA, Wang L, Bunker CH, Lowder JL. Episiotomy in the United States: has anything changed?. Am J Obstet Gynecol 2009;200:573.e1–7.
6. Dennison C, Pokras R. Design and operation of the National Hospital Discharge Survey: 1988 redesign. Vital Health Stat 1 2000;1:1–42.
8. Marjoribanks J, Lethaby A, Farquhar C. Surgery versus medical therapy for heavy menstrual bleeding. The Cochrane Database of Systematic Reviews 2006, Issue 2. Art. No.: CD003855. DOI: 10.1002/14651858.CD003855.pub2.
9. Lowder JL, Oliphant SS, Ghetti C, Burrows LJ, Meyn LA, Balk J. Prophylactic bilateral oophorectomy or removal of remaining ovary at the time of hysterectomy in the United States, 1979–2004. Am J Obstet Gynecol 2010;202:538.e1–9. Epub 2010 Jan 13.
10. Lepine LA, Hillis SD, Marchbanks PA, Koonin LM, Morrow B, Kieke BA, Wilcox LS. Hysterectomy surveillance-United States, 1980–1993. MMWR CDC Surveill Summ 1997;46:1–15.
11. Luber KM, Boero S, Choe JY. The demographics of pelvic floor disorders: Current observations and future projections. Am J Obstet Gynecol 2001;184:1496–503.
12. Jones KA, Shepherd, JP, Oliphant SS, Wang L, Bunker CH, Lowder JL. Trends in inpatient prolapse procedures in the United States, 1979–2006. Am J Obstet Gynecol 2010;202:501.e1–7. Epub 2009 Mar 11.
13. Menacker F, Martin JA. BirthStats: rates of cesarean delivery, and unassisted and assisted vaginal delivery, United States, 1996, 2000, and 2006. Birth 2009;36:167.
14. Frankman EA, Wang L, Bunker CH, Lowder JL. Episiotomy in the United States: has anything changed? Am J Obstet Gynecol 2009;200:573.e1–7. Epub 2009 Feb 24.
15. Carroli G, Belizan J. Episiotomy for vaginal birth. The Cochrane Database of Systematic Reviews 2000, Issue 2. Art. No.: CD000081. DOI: 10.1002/14651858.CD000081.
17. American Medical Association. Physician characteristics and distribution in the US. American Medical Association. 1997/8–2008 editions. Chicago (IL): American Medical Association.
© 2010 by The American College of Obstetricians and Gynecologists.