Cesarean delivery is usually considered a safe, low-risk procedure by most health care practitioners and the general public. Although vaginal birth after previous cesarean (VBAC) rates decreased from 35.3% in 1994–1995 to 29.9% in 2000–2001, overall cesarean delivery rates in Canada increased from 17.5% to 21.2% over the same period.1 Canada now has one of the highest rates of cesarean delivery among industrialized countries.1–4 In the United States, cesarean delivery rates increased from 20.7% in 1996 to 22.5% in 2000.5 Previous studies on the benefits and risks of this mode of delivery have focused on immediate outcomes.6–9 Because some elective cesarean deliveries are now being performed for nonmedical reasons, better and more complete information is required, especially on longer-term postpartum outcomes.
Maternal readmission is one indicator of postpartum maternal morbidity.1,10–12 Our previous study of maternal readmission focused on the association of initial length of in-hospital stay with obstetric conditions at child birth, and with maternal readmission.13 That study revealed that maternal postpartum readmission was associated with initial length of in-hospital stay independent of mode of delivery. More information about the safety of cesarean delivery is warranted, especially given that a growing number of women are requesting elective cesarean delivery in recent years (Hall MH, Bewley S. Maternal mortality and mode of delivery [letter]. Lancet 1999;354:776).6,14–16 Also, trends and patterns of cesarean delivery have changed substantially in recent years.4,17 We therefore studied the relationship between delivery method and postpartum readmission within 60 days after discharge from the child birth hospitalization, using information on hospital deliveries from Canadian provinces and territories over the period from 1997 to 2001.
MATERIALS AND METHODS
The Canadian Institute for Health Information began collecting information on all admissions to most of Canada's acute-care hospitals in the early 1980s. After assessment of its potential uses and limitations,18,19 Canadian Institute for Health Information's hospital Discharge Abstract Database has been widely used for perinatal surveillance and research.1,12,13,17 Obstetric deliveries in the Discharge Abstract Database are identified using case-mix group, diagnosis, procedure, and patient service (ie, obstetric) codes.1 Hospital deliveries occurring from April 1, 1997, through March 31, 2001, were included in the present study. Deliveries occurring in Quebec and Manitoba were not included, because complete information on these provinces is not contained in the Discharge Abstract Database. The total number of deliveries in the study provinces/territories was 957,998, which accounts for approximately 98% of all deliveries that occurred during the 4-year period (excluding the approximately 2% of deliveries occurring at home or in birthing centers). Data available from these hospital discharge records include health insurance number, demographic and residence information, date of admission, date and status at discharge, principal diagnosis and up to 15 secondary diagnoses (coded according to the International Classification of Diseases, 9th Revision, Clinical Modification [ICD-9-CM]), and up to 10 diagnostic, therapeutic, and surgical procedures (coded according to the Canadian Classification of Diagnostic, Therapeutic, and Surgical Procedures [CCP]).
Obstetric deliveries were excluded if maternal age at admission for childbirth was less than 15 years or more than 44 years, if the deliveries resulted in multiple birth or fetal death or in-hospital infant death, if a discharge occurred on the same day of admission, if the mother remained in the delivery hospital for more than 20 days or was transferred within the same hospital or to another hospital during the delivery hospitalization, or if she had any of the following conditions: diabetes mellitus, pre-existing hypertension, cardiac or renal disease, multiple sclerosis, or systemic lupus erythematosus. A total of 900,108 obstetric deliveries remained in the final analysis, which represents 94.0% of all hospital deliveries during the study period.
In the Discharge Abstract Database, all records have a completely unique health insurance number assigned by the provincial/territorial health authority. These health insurance numbers (scrambled by a systematic algorithm) were used to identify readmission cases by deterministic record linkage (ie, records of hospital deliveries were linked to any subsequent records of hospital admission within a given period). The validity of each linkage was confirmed by further matching the mother's date of birth on the 2 records and/or postal code of residence. Readmission in this study was defined as a hospitalization that occurred within 60 days after the initial discharge from the delivery hospitalization. Day surgery after discharge after delivery was not considered as a readmission. Similarly, rehospitalization for elective procedures such as sterilization was not included in readmission counts.
The CCP codes on the hospital obstetric records were used to determine cesarean deliveries (CCP 86.0, 86.1, 86.2, 86.8, 86.9), forceps vaginal deliveries (CCP 84.0, 84.1, 84.2, 84.3, 84.4, 84.6), or vacuum vaginal deliveries (CCP 84.7). Cesarean deliveries were further subdivided into emergency cesarean and elective cesarean deliveries according to previously specified CCP and ICD-9 codes.4 We examined maternal complications at initial hospital discharge by method of delivery, including premature rupture of membranes, postpartum hemorrhage, preterm labor, breech presentation, preeclampsia/eclampsia, abruptio placenta, and total length of stay for childbirth. We also determined the frequency of each principal diagnosis for all readmitted cases. Such readmission diagnoses were classified into major categories that included genitourinary conditions, obstetric surgical complications, postpartum hemorrhage, major puerperal infection, pelvic injury/wounds, venous disorders and thromboembolism, psychiatric disorders, cardiopulmonary conditions, gastrointestinal conditions, gallbladder diseases, and breast infections. We assessed the association between method of delivery and subsequent postpartum readmission (as well as diagnosis-specific postpartum readmission) using multiple logistic regression. We used separate logistic models for each outcome, adjusting for maternal age, based on the assumption that medical conditions for readmission occur independently. Odds ratios (ORs) and 95% confidence intervals (CIs) were calculated using spontaneous vaginal delivery as the reference. Adjusted rates were used to calculate the number of cesarean, forceps, or vacuum deliveries that would be required to produce 1 excess postpartum readmission (compared with spontaneous vaginal delivery). Fisher exact test was used to test differences in in-hospital mortality during rehospitalization between different methods of delivery. We also examined additional information about maternal readmission pertaining to delivery method, including the interval between previous discharge and readmission, length of hospital stay, and distribution of the principal diagnosis among readmitted women.
Maternal age and several obstetrical complications at delivery hospitalization were associated with method of delivery (Table 1). The proportions of women 30–34 years and 35 years or older were highest in the cesarean delivery group. Postpartum hemorrhage was more frequently observed in women with operative or spontaneous vaginal delivery, whereas breech presentation, preterm labor, preeclampsia/eclampsia, abruptio placenta, chorioamnionitis, and placenta previa were more common in the cesarean delivery group. As expected, women with cesarean delivery remained in hospital significantly longer than those delivered vaginally (Table 1).
A total of 16,404 women were readmitted within 60 days after discharge after the delivery hospitalization, resulting in an overall readmission rate of 1.8%. Women with cesarean delivery, forceps vaginal delivery, and vacuum vaginal delivery had readmission rates of 2.7%, 2.2%, and 1.8%, respectively, compared with 1.5% of women with spontaneous vaginal delivery (P <.001).
After adjusting for maternal age, women undergoing cesarean delivery were at significantly elevated risk of postpartum readmission (OR 1.87, 95% CI 1.80–1.93); ie, there was 1 excess postpartum readmission per 75 cesarean deliveries, relative to spontaneous vaginal delivery. Diagnoses associated with this increased risk included pelvic injury/wounds (OR 13.38, 95% CI 11.96–14.98), obstetric surgical complications (OR 3.01, 95% CI 2.63–3.45), venous disorders and thromboembolism (OR 2.71, 95% CI 2.13–3.44), and major puerperal infection (OR 1.76, 95% CI 1.61–1.92). Nonobstetric diagnoses such as cardiopulmonary, gastrointestinal, genitourinary and psychiatric disorders also occurred more frequently among readmitted women after cesarean delivery (Table 2).
Women delivered by forceps were also at an increased risk of postpartum readmission (OR 1.43, 95% CI 1.34–1.53); ie, there was 1 excess readmission per 152 forceps vaginal deliveries (compared with spontaneous vaginal delivery). Diagnoses associated with this increased risk of readmission were pelvic injury/wounds (OR 5.56, 95% CI 4.65–6.64), obstetric surgical complications (OR 1.71, 95% CI 1.31–2.22), major puerperal infection (OR 1.62, 95% CI 1.40–1.87), breast infections (OR 1.43, 95% CI 1.11–1.85), and postpartum hemorrhage (OR 1.19, 95% CI 1.04–1.37). In addition, women with delivered by forceps were at increased risk of several nonobstetric disorders (eg, gallbladder), compared with women who had spontaneous vaginal delivery.
Women who had a vacuum delivery were at a slightly increased risk of postpartum readmission (OR 1.21, 95% CI 1.15–1.28); ie, there was 1 excess readmission per 311 vacuum deliveries (compared with spontaneous vaginal delivery). Diagnoses associated with the increased risk of readmission were similar to those with forceps vaginal delivery, although the associations were weaker (Table 2).
Of the readmitted women, 13 died in hospital, with in-hospital mortality rates of 2.83 (5 cases), 0, 1.25 (1 case), and 1.18 (7 cases) per 100,000 deliveries for cesarean, forceps, vacuum, and spontaneous vaginal routes, respectively. Compared with spontaneous vaginal delivery, no statistically significant increase in mortality was observed either in the cesarean delivery group (P = .12) or in forceps (P = .57) and vacuum delivery groups (P = .63; Table 2).
Women with different modes of delivery differed in average length of in-hospital stay for childbirth, interval between initial discharge and postpartum readmission, and length of in-hospital stay after rehospitalization. On average, women who had a cesarean delivery were rehospitalized much sooner than those who had a spontaneous vaginal delivery (14.9 versus 19.7 days), and also sooner than those with delivery by forceps (15.8 days). The length of hospital stay after rehospitalization averaged 3.5 days for cesarean, 2.9 days for forceps, and 2.7 days for both vacuum and spontaneous vaginal delivery (Table 3).
Postpartum hemorrhage was the most commonly observed principal diagnosis for readmission among rehospitalized women with forceps, vacuum, or spontaneous vaginal delivery (19.8%, 23.6% and 23.6% of readmitted women, respectively). Puerperal complications were the most common cause for readmission among women rehospitalized after cesarean delivery (26.8%). Major puerperal infection was the second principal reason for readmission for all modes of delivery (13.2–16.2%). In addition, gallbladder disorders were the most common nonobstetric cause of postpartum readmission for all modes of delivery (7.5–10.2%).
Readmission rates were 2.9% and 2.6%, respectively, after emergency cesarean delivery and elective cesarean delivery. Odds ratios were 1.95 (95% CI 1.86–2.05, P < .001) and 1.78 (95% CI 1.71–1.87), respectively, for emergency cesarean and elective cesarean delivery compared with spontaneous vaginal delivery. The associations between emergency/elective cesarean delivery and specific readmission diagnoses were of similar magnitude. The ORs for emergency cesarean and elective cesarean delivery and pelvic injury/wounds were 16.06 (95% CI 14.22–18.14) and 11.16 (95% CI 9.83–12.68), respectively (relative to spontaneous vaginal delivery).
This population-based cohort study examined women's risk of postpartum readmission associated with mode of delivery. Cesarean delivery and operative (forceps or vacuum extraction) vaginal delivery were associated with a significantly elevated risk of maternal postpartum readmission for several serious obstetric conditions (eg, pelvic injury/wounds, obstetric surgical complications, major puerperal infection, and venous disorders and thromboembolism). We also observed that women were readmitted to hospital significantly sooner after cesarean delivery, and stayed in hospital longer, compared with women after spontaneous vaginal delivery.
Medical attention is usually focused on the immediate events after an obstetric delivery.6–9,16,17 Obstetricians and women typically attempt to balance shorter-term risks and benefits when deciding on the mode of delivery. Information on maternal postpartum readmission and its principal reasons, associated complications, and length of in-hospital stay after rehospitalization has been infrequently reported in the literature.10,13,20,21 Because readmission is often associated with increased suffering, higher costs, disruption of early parenting, and increased family burden, it is an important indicator for monitoring longer-term postpartum maternal health status that is difficult for obstetric care providers to anticipate at the time of childbirth.1,12 Our study shows that cesarean delivery poses a higher risk of readmission than those with other modes of delivery. Obstetricians and women should be aware of the potential increase in risk when contemplating choice of delivery mode.
Lydon-Rochelle et al10 analyzed data for Washington State and found an overall readmission rate of 1.2% within 60 days of delivery; readmission was nearly twice as common after cesarean delivery as after spontaneous vaginal delivery. Similarly, our data showed that compared with women who had a spontaneous vaginal delivery (1.5 readmissions per 100 deliveries), women who had a cesarean delivery had an 87% increased risk of postpartum readmission (2.7 per 100 deliveries), women delivered by forceps had a 43% increased risk of postpartum readmission (2.2 per 100 deliveries), and those delivered by vacuum had a 21% increased risk of postpartum readmission (1.8 per 100 deliveries) within 60 days after discharge from the delivery hospitalization.
After cesarean delivery, women had a significantly elevated risk of rehospitalization as a result of a number of obstetric complications. The most important of these was a 13-fold increase in risk of pelvic injury or wound infection among women with cesarean delivery. The rate of rehospitalization for this condition was approximately 9 per 1,000 cesarean deliveries. It is expected that a higher risk of postpartum readmission for surgical wound infection would occur among women with cesarean delivery, because all such women have a surgical wound, whereas most women with spontaneous vaginal delivery do not even undergo an episiotomy. Women with operative vaginal delivery also showed a higher risk of rehospitalization due to pelvic injury/wounds (2–4 per 1,000 deliveries), however, suggesting that the degree of trauma incurred is sometimes severe enough to lead to postpartum rehospitalization.
Our study confirms previous findings10,20–24 that obstetric surgical procedures, including cesarean and forceps or vacuum extraction, increase the risk of postpartum rehospitalization. Other studies examining women during childbirth hospitalization or postpartum rehospitalization have also reported an increased risk of obstetric and puerperal complications and of cardiopulmonary, thromboembolic, gastrointestinal, and gallbladder conditions associated with cesarean delivery or operative vaginal delivery.17,23–26 Undoubtedly, some of these conditions may be caused or exacerbated by a surgical procedure or wound infection. It has been reported that an increased risk of rehospitalization for appendicitis among women after cesarean delivery is associated with the manipulation of abdominal contents during the operation.27,28 Thus, it is not surprising that postpartum morbidity is more common in women with an operative delivery. However, it is also possible that the excess risk of readmission for severe conditions, such as gallbladder disorders, may have been due to factors that increased the risk of cesarean delivery. For example, obese women who are more likely to be delivered by the cesarean route are also more likely to experience cholelithiasis and other gallbladder disorders.
As expected, the length of initial hospital stay for childbirth varies significantly by delivery method. The interval between discharge after delivery hospitalization and readmission and the length of in-hospital stay after readmission are also both associated with method of delivery. These indices may serve as a proxy for maternal postpartum conditions for which hospitalization is required. Many women were readmitted after cesarean delivery with a primary diagnosis of puerperal complications (27.0%) or major puerperal infection (13.2%), whereas for women delivered vaginally, postpartum hemorrhage was the most frequent reason (19.8–23.6%), followed by major puerperal infection (13.8–16.2%). These findings suggest that women should be carefully assessed for signs and symptoms of infection, hemorrhage, and other puerperal complications before discharge after the delivery event. Women at risk for developing these complications after delivery hospitalization discharge should receive appropriate follow-up in the community.
Several limitations of our study merit discussion. The associations we observed may have been confounded by unmeasured factors. The increased risk of maternal postpartum rehospitalization after cesarean or operative vaginal delivery may be partially attributable to factors other than delivery method (ie, to factors that affected the choice of method of delivery). However, pregnant women with severe pre-existing or chronic medical conditions that could have predisposed them both to cesarean or operative delivery and to rehospitalization were excluded from our analysis. In practice, maternal rehospitalization may be influenced by socioeconomic and demographic factors and health care accessibility. Parity is one important factor associated both with choice of method of delivery and with maternal rehospitalization. Unfortunately, information on parity was not available in the study database. Finally, administrative data are prone to a certain degree of coding error and incompleteness.29,30 Coding errors are likely to have occurred randomly, which would tend to attenuate the effects observed.
In summary, the results of this large population-based cohort study suggest that operative vaginal delivery and cesarean delivery are associated with a higher risk of maternal postpartum readmission. This increased risk is due to an increased frequency of postpartum readmission for several serious complications (eg, pelvic injury and wound infection). These results add a further dimension of information that should help obstetricians and women when discussing the benefits and risks associated with spontaneous vaginal delivery, operative vaginal delivery, and cesarean delivery.
1. Health Canada. Canadian Perinatal Health Report 2003. Ottawa: Ministry of Public Works and Government Services Canada; 2003.
2. Dobson R. Caesarean section rate in England and Wales hits 21%. BMJ 2001;323:951.
3. Martin JA, Hamilton BE, Ventura SJ, Menacker F, Park MM. Births: final data for 2000. Natl Vital Stat Rep 2002;50(5):1–101.
4. Liu S, Rusen ID, Joseph KS, Liston R, Kramer MS, Wen SW, et al, for the Maternal Health Study Group of the Canadian Perinatal Surveillance System. Recent trends in caesarean delivery rates and indications for caesarean delivery in Canada. J Obstet Gynaecol Can 2004;25:735–42.
5. American College of Obstetricians and Gynecologists. Women's Health Care Physicians, Task Force on Cesarean Delivery Rates. Evaluation of cesarean delivery. Washington, DC: American College of Obstetricians and Gynecologists; 2000.
6. Cnattingius R, Cnattingius S, Notzon FC. Obstacles to reducing cesarean rates in a low-cesarean setting: the effect of maternal age, height, and weight. Obstet Gynecol 1998;92:501–6.
7. Hannah ME, Hannah WJ, Hewson SA, Hodnett ED, Saigal S, Willan AF. Planned cesarean section versus planned vaginal birth for breech presentation at term: a randomized multicentre trial. Term Breech Trial Collaborative Group. Lancet 2000;356:1375–83.
8. Hillian EM. Postoperative morbidity following cesarean delivery. J Adv Nurs 1995;22:1035–42.
9. Petitti DB. Maternal mortality and morbidity in cesarean section. Clin Obstet Gynecol 1985;28:764–9.
10. Lydon-Rochelle M, Holt VI, Martin DP, Easterling TR. Association between method of delivery and maternal rehospitalization. JAMA 2000;283:2411–6.
11. Glazener CM, Abdalla M, Stroud P, Naji S, Templeton A, Russell IT. Postnatal maternal morbidity: extent, causes, prevention and treatment. Br J Obstet Gynaecol 1995;102:282–7.
12. Health Canada Canadian Perinatal Health Report 2000. Ottawa: Ministry of Public Works and Government Services Canada; 2000.
13. Liu S, Heaman M, Kramer MS, Demissie K, Wen SW, Sylvie Marcoux. Length of hospital stay, obstetric conditions at childbirth, and maternal readmission: a population-based cohort study. Am J Obstet Gynecol 2002;187:681–7.
14. Hannah ME. Planned elective cesarean section: a reasonable choice for some women? CMAJ 2004;170:813–4.
15. Surgery and patient choice: the ethics of decision making ACOG Committee Opinion No. 289. American College of Obstetricians and Gynecologists. Obstet Gynecol 2003;102:1101–6.
16. Gilbert MG, Hicks SM, Boe NM, Danielsen B. Vaginal versus cesarean delivery for breech presentation in California: a population-based study. Obstet Gynecol 2003;102:911–7.
17. Wen SW, Rusen ID, Walker M, Liston R, Kramer MS, Baskett T, et al, for the Maternal Health Study Group, Canadian Perinatal Surveillance System. Comparison of maternal mortality and morbidity between trial of labor and elective cesarean section among women with previous cesarean delivery. Am J Obstet Gynecol 2004;191:1263–9.
18. Wen SW, Liu S, Marcoux S, Fowler D. Uses and limitation of routine hospital admission/separation records for perinatal surveillance. Chronic Dis Can 1997;18:113–9.
19. Wen SW, Liu S, Marcoux S, Fowler D. Trends and variations in length of hospital stay for childbirth in Canada. CMAJ 1998;158:875–80.
20. Couto RC, Pedrosa TMG, Nogueira JM, Gomes DL, Neto MF, Rezende NA. Post-discharge surveillance and infection rates in obstetric patients. Int J Gynaecol Obstet 1998;61:227–31.
21. Vermillion ST, Lamoutte C, Soper DE, Verdeja A. Wound infection after cesarean: effect of subcutaneous tissue thickness. Obstet Gynecol 2000;95:923–5.
22. Kolderup LB, Laros RK, Musci TJ. Incidence of persistent birth injury in macrosomic infants: association with mode of delivery. Am J Obstet Gynecol 1997;177:37–41.
23. Nielsen TF, Hokegard KH. Postoperative cesarean section morbidity: a prospective study. Am J Obstet Gynecol 1983;146:911–6.
24. Martens MG, Kolrud BL, Faro S, Maccato M, Hammill H. Development of wound infection or separation after cesarean delivery: prospective evaluation of 2,431 cases. J Reprod Med 1995;40:171–5.
25. Leigh DA, Emmanuel FX, Sedgwick J, Dean R. Postoperative urinary tract infection and wound infection in women undergoing caesarean section. J Hosp Infect 1990;15:105–6.
26. Boggess KS, Watts DH, Hillier SL, Krohn MA, Benedetti TJ, Eschenbach DA. Bacteremia shortly after placental separation during cesarean delivery. Obstet Gynecol 1996;87:779–84.
27. Anderson R, Hugander A, Thulin A, Nystrom P, Olaison G. Clusters of acute appendicitis: further evidence for an infectious aetiology. Int J Epidemiol 1995;24:829–33.
28. Hemminki E. Long term maternal health effects of caesarean section. J Epidemiol Comm Health 1991;45:24–8.
29. Huston P, Naylor CD. Health services research: reporting on studies using secondary data sources. CMAJ 1996;155:1697–702.
30. Iezzoni LI, Foley SM, Daley J, Hughes J, Fisher ES, Heeren T. Comorbidities, complications and coding bias: does the number of diagnosis codes matter in predicting in-hospital mortality? JAMA 1992;267:2197–203.
This study was carried out under the auspices of the Canadian Perinatal Surveillance System. Contributing members of the Maternal Health Study Group include Drs. Tom Baskett (Dalhousie University), Leanne Dahlgren (University of British Columbia), Susie Dzakpasu (Public Health Agency of Canada), William Fraser (University of Montreal), Robert Kinch (McGill University), Catherine McCourt (Public Health Agency of Canada), Hajnal Molnar-Szakács (Public Health Agency of Canada), I. D. Rusen (Public Health Agency of Canada), and Shi Wu Wen (University of Ottawa).