| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
ORIGINAL RESEARCH |
From the School of Health Care Administration, Taipei Medical University, Taipei, Taiwan; and Arnold School of Public Health, University of South Carolina, Department of Health Services Policy and Management, Columbia, South Carolina.
Address reprint requests to: Herng-Ching Lin, School of Health Care Administration, Taipei Medical University, 250 Wu-Hsing Street, Taipei 110, Taiwan; e-mail: henry11111{at}tmu.edu.tw.
 |
ABSTRACT |
|---|
TOP
 ABSTRACT MATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
METHODS: Cross-sectional data from Taiwan’s National Health Insurance database was used, covering all 270,774 women admitted for singleton deliveries, in 2000. Bivariate and multiple logistic regression analyses were used.
RESULTS: The overall cesarean rate was 32.3% of all deliveries. Obstetrics and gynecology clinics (with fewer than 10 beds) had a very high likelihood of cesarean delivery compared with all categories of hospitals (odds ratios 17–25), after adjusting for clinical complications and patient, physician, and institutional characteristics. The likelihood of cesarean delivery was similar across hospitals, regardless of level and ownership category. High cesarean propensity at clinics arose from higher cesarean rates in all complication categories, including "No complications." The overall hospital cesarean rate, 31.2%, is also higher than that in other developed countries with universal health care coverage.
CONCLUSION: Taiwan has very high cesarean rates, with a particularly high propensity for this procedure at clinics. The cesarean delivery profile in the various clinical complication categories suggests a significantly lower clinical threshold triggering cesarean delivery decisions in Taiwan, especially at obstetrics and gynecology clinics. Countries currently having or contemplating large expansions in health insurance coverage should document obstetric practice profiles before initiating coverage expansions. There is also a need for well designed research on the medical and life-satisfaction impacts of cesarean compared with vaginal delivery to enable an informed policy stand on this issue.
LEVEL OF EVIDENCE: III
Inappropriate cesarean deliveries increase maternal and neonatal morbidity and health care costs.2,9 Often, nonclinical factors play a crucial role in elective cesarean delivery, as demonstrated in Brazil.10 Internationally, researchers have demonstrated the role of patient factors, such as socioeconomic status,11 race,12 type of insurance,13 provider characteristics, such as practice styles,14 age,15 gender,15 delivery in daylight hours,16 convenience factors,16 fear of litigation,17 and type of birth attendants.14 Institutional factors are also documented, namely, hospital size,15 teaching status,18 and ownership.18 To date, most research has relied on regional samples, samples from selected hospitals or patient subpopulations (eg, those covered by a specific insurance plan), or samples lacking the required clinical information. The use of national data sets with comprehensive clinical information across all providers and deliveries circumvents selection bias and confounding and can help formulate effective policies to steer the health system toward appropriate obstetric care practices that are consistent with the clinical profile of cases.
The National Health Insurance Database of Taiwan presents a unique opportunity to identify institutional factors affecting cesarean delivery rates, adjusted for clinical and demographic factors. Taiwan has universal health insurance, a single-payer system with government as the sole insurer and payer, comprehensive benefits, unrestricted access to any provider, and a variety of public and private providers competing with one another for patients. Reimbursement rates for cesarean and vaginal delivery are fixed (regardless of resource use), with the rate for cesarean delivery being twice that of vaginal delivery. Most pregnant women remain with one provider through the antenatal period until delivery. The national database, covering virtually every delivery in Taiwan, enables systematic exploration of institutional clinical behavior under different forms of ownership and levels of clinical capability. Unlike many earlier studies, the strength of this study lies in its statistical power to isolate the confounding effects of fairly uncommon clinical complications, enabling more robust conclusions.
|
MATERIALS AND METHODS |
|---|
|
TOP
ABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The dependent variable was dichotomous, whether or not a cesarean delivery was performed. The key independent variables were institutional ownership (public, voluntary nonprofit, and private for-profit), geographic location (north, central, south, and east Taiwan), teaching status (teaching or nonteaching institution), and hospital level. Under National Health Insurance, hospitals are classified into 4 levels (based on bed capacity and clinical capabilities): medical centers (minimum 500 beds), regional hospitals (minimum 250 beds), district hospitals (minimum 20 beds), and clinics (fewer than 10 beds). All medical centers and regional hospitals are teaching hospitals, as are some district hospitals. Total institutional caseload could also affect cesarean rates and is operationalized as "Size of delivery service," in line with earlier studies.19
To isolate the effect of institutional ownership, while differentiating between obstetrics and gynecology clinics and hospitals, additional analyses were done, classifying all institutions into 4 categories: public hospitals, private hospitals, voluntary hospitals, and obstetrics and gynecology clinics (fewer than 10 beds, all privately owned). Private clinics have qualitatively different dynamics in place, being almost always managed by solo practitioners and exempt from the hospital accreditation requirements of the Bureau of National Health Insurance. Hospital teaching status was included in the regressions where appropriate.
Other factors influencing delivery type are physician variables, patient-related variables, and clinical indications. We controlled for physician’s gender, age (a surrogate for duration of practice experience), and patient’s age. Patient parity is not available in this data set. All clinical complications (secondary diagnoses) were classified according to a standard hierarchy of mutually exclusive diagnoses devised by Anderson and Lomas20 and used by several authors.21,22 All deliveries were assigned to one of the following categories: 1) previous cesarean (ICD-9-CM 654.2), 2) breech presentation (652.2 and 669.6), 3) dystocia (653 and 660–662, excluding 661.3), 4) fetal distress (656.3), 5) other antepartum or intrapartum complications potentially justifying a cesarean, 6) complications reflecting pelvic floor/perineal/birth canal injury sustained during vaginal delivery, 7) other comorbidities not ordinarily indications for cesarean, and 8) no complications (no secondary diagnosis). The first 4 conditions form Anderson’s and Lomas’s20 hierarchy in that order, so that any case with 2 or more complications is allocated to the complication that takes precedence over the others. For example, a patient with dystocia and previous cesarean delivery is classified to the latter category, because it stands higher in Anderson’s hierarchical ordering.
After classifying all deliveries as above, 3 categories—breech presentation, dystocia, and fetal distress—were collapsed into a single category, titled "Unequivocal indications for cesarean," keeping in view their higher expectancy for a cesarean delivery according to the current state of the art.23 We retained "Previous cesarean" as a separate category because the professional consensus on trial of labor varies greatly, as reflected in the range of 45% vaginal delivery rates in Denmark to 7% in the United States.22,24 Therefore, 6 clinical categories were used in the regression analysis: 1) previous cesarean, 2) unequivocal indications for cesarean, 3) other complications potentially justifying cesarean, 4) pelvic floor/perineal/birth canal injuries, 5) Other comorbidities not ordinarily indications for cesarean, and 6) no complications. The diagnoses classified under categories 3, 4, and 5 are listed in the Appendix.
|
RESULTS |
|---|
|
TOP
ABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
shows the univariate statistics by delivery type, hospital level and ownership, clinical complications, and attending physician’s characteristics. The overall cesarean delivery rate was 32.3% of all deliveries, and mean patient age was 28 years (standard deviation ± 5 years).
|
) show that medical centers, public hospitals, southern institutions, and teaching hospitals had higher cesarean delivery rates than did the remaining categories in the respective classifications. However, the magnitudes of the differences were marginal, and all findings were reversed after adjusting for clinical, demographic, and provider variables.
|
shows the bivariate statistics of hospital level by clinical complication, patient’s age, and attending physician’s age. Medical centers and regional hospitals had the highest rate of all clinical complications, followed by district hospitals and obstetrics and gynecology clinics (
2 = 39,243, P < .001).
|
shows the adjusted odds of cesarean delivery. Compared with medical centers, the likelihood of cesarean delivery at obstetrics and gynecology clinics was 20-fold greater (reciprocal of 0.05); compared with regional hospitals it was 25-fold, and compared with district hospitals, 17-fold, after adjusting for other institutional variables, clinical complications, patient’s age, and physician characteristics. In this regression, teaching status was excluded to circumvent collinearity, because all medical centers and regional hospitals are teaching hospitals, and clinics are not. Apart from the outstanding likelihood of cesarean delivery at obstetrics and gynecology clinics, the odds ratios were similar among the different levels of hospitals. The influence of ownership is marginal: private institutions were 1.4 times as likely as public hospitals and twice as likely as voluntary hospitals to have cesarean deliveries.
|
).
The clinical categories "previous cesarean," "unequivocal indications for cesarean," "other complications potentially justifying cesarean," and "other comorbidities not ordinarily indications for cesarean" have very high likelihood of cesarean delivery relative to uncomplicated deliveries (odds ratios greater than 999; Table 4). As expected, "complications reflecting birth canal injuries" almost always supervene on a vaginal delivery and showed lower odds of cesarean delivery compared with "no complications." Older women (aged more than 35 years) were more likely to have cesarean delivery, in line with clinical expectations. Physician’s age and gender were not significantly associated with cesarean delivery.
We explored complication-specific cesarean rates by hospital level (Table 5). In all complication categories except birth canal injuries, clinics had the highest cesarean rates, and medical centers/regional hospitals had the lowest rates. Overall cesarean rates were very high in the categories of "previous cesarean" (98.1%, range 95.4–99.8%), and "unequivocal indications for cesarean" (95.1%, range 88.9–99.8%), with clinics having the highest rates in both categories.
|
|
DISCUSSION |
|---|
|
TOP
ABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The higher likelihood of cesarean delivery at clinics is also attributable to physicians’ time pressures. Clinics in Taiwan are mostly run by solo practitioners working, on average, 9.36 hours a day, 6.2 days a week.26 Solo obstetricians are more vulnerable to overwork, often having to wait out unpredictable hours of labor. A scheduled cesarean delivery is a potential time-management solution, especially if comorbidities require close obstetric monitoring during labor. Our finding of disproportionately higher cesarean rates at clinics in all clinical categories supports this explanation (Table 5). Even the category "other comorbidities not ordinarily indications for cesarean" had a 90.7% rate at clinics compared with 24.8% at medical centers and 16.7% at regional hospitals. Clearly, clinics in Taiwan have a very low clinical threshold triggering a cesarean delivery decision. Importantly, in our study, we adjusted for physician’s age, a reasonable proxy for duration of practice experience. Therefore, the excess cesarean likelihood at clinics is not attributable to inadequate clinical experience. Other authors have also suggested that physicians in Brazil may schedule cesarean deliveries to shorten their working hours.3
Another important factor is the differential regulatory treatment of hospitals. All hospitals have to meet Bureau of National Health Insurance accreditation requirements (suggesting a 30% ceiling on cesarean rates) and periodic scrutiny. Although it is not strictly enforced, hospitals have internal reviews that ensure physicians stay close to the norm to avoid attracting the negative attention of the Bureau. The regulatory effect may also explain why cesarean rates are quite similar in private, public, and voluntary hospitals. Clinics, being exempt from accreditation requirements, have less pressure to re-evaluate their clinical decisions.
Despite the clinics’ high propensity for cesarean delivery and the fact that they handle 34.8% of all deliveries, their net excess contribution to the total cesarean volume in Taiwan remains small (2.7% of all cesareans). This is due to the substantially lower prevalence of complications among clinic patients and their favorable maternal-age profile for vaginal delivery (Table 3).
Taiwan’s net cesarean rate of 31.2% of hospital deliveries, (excluding clinic deliveries) is also a matter of concern. Nationwide, cesarean rates have hovered between 32.3% and 34% since National Health Insurance implementation. Although nationwide, pre-National Health Insurance information is not available, post-National Health Insurance rates probably represent a big jump. Tsai et al demonstrated that before National Health Insurance, uninsured women had a cesarean rate of 8.3%. Under National Health Insurance, this rate jumped to 33% at one large nonprofit hospital system in Taiwan.27 Their study covered 11,788 primiparas, which is statistically inadequate to adjust for clinical complications. Our study, including all birth orders across all medical institutions and covering 270,774 deliveries, showed a rate of 32.3%, similar to Tsai’s finding and substantiating it on a nationwide scale without selection biases and confounding. Taiwan’s rate could be also be partly driven by patient preferences for astrologically auspicious delivery timings, convenience of a scheduled delivery, avoidance of delivery pain, concerns about sexual functioning following vaginal delivery, and insurance coverage, as documented by Huang et al (527 women; response rate 92.7%).27
Taiwan’s rate is much higher than England’s 21.4%,4 Norway’s 12.8%, and Sweden’s 10.8%,23 all countries with universal health coverage. The pediatric and maternal health trade-offs from Taiwan’s higher cesarean rate remain dubious. A crude indicator, infant mortality rate, has remained steady, about 6 per 1000 live births, before and since the implementation of National Health Insurance, which is comparable to the infant mortality rate in these countries. Moreover, in Taiwan, one cesarean delivery precludes future vaginal deliveries, as demonstrated by the 98.1% cesarean rate among previous cesarean cases, compared with Denmark’s 45%, Hungary’s 68%, and the United States’ 93%.22,24 Taiwan also has disproportionately higher cesarean rates among breech, dystocia, and fetal distress cases (95%), compared with 35% among breech deliveries in the Netherlands, 69% in Canada, 80% in the United States and the former West Germany, and 93% in Sweden.24 In dystocia cases, the United States had a 65% cesarean rate, and in fetal distress cases, 63%.24
Disproportionately high cesarean rates in Taiwan hold major lessons for the many countries contemplating or having universal health insurance coverage, with a similar mix of providers. Governments should be aware of the remarkable potential for an increase in cesarean rates. Unlike Taiwan, they should proactively document baseline clinical practice profiles and the associated maternal and fetal outcomes before initiating major coverage expansion. This will help establish optimum practice benchmarks adjusted for outcomes, which in turn could be used as institutional accreditation guidelines. Countries with large or universally insured populations should evaluate delivery profiles in relation to institutional size and reimbursement policies.
At this point, our findings on clinics raise some critical issues. Our research suggests that accreditation helps maintain practice profiles closer to the expected norms. Accreditation is an expensive process and is driven by statistical assessments. But low delivery volumes at individual clinics preclude the monitoring of practice and outcomes, adjusted for complication rates. Infrastructure accreditation alone without clinical profiling will be of questionable value, besides being prohibitively expensive (given the size and number of clinics). Should insurance coverage be limited to hospitals with delivery volumes amenable to clinical profiling? Restricting the scope of clinics to outpatient and ambulatory care may not be acceptable to a profession that is largely dependent on deliveries for its practice clientele. Hard choices have to be made, and we hope that research studies such as this one can help in making informed policy choices.
A fundamental issue raised by our study (and others in the past) relates to the mix of myth and reality driving maternal and obstetricians’ delivery preferences. Some obstetricians prefer liberal cesarean policies, citing pelvic floor damage, fetal injuries, and patient concerns about sexual functioning following compromised vaginal deliveries.4,28 Other obstetricians cite the adverse maternal and pediatric sequelae of cesarean delivery ( van Roosmalen J. Unnecessary cesarean sections should be avoided [letter]. BMJ 1999;318:121.).25 In our study, the incidence of birth canal injuries was 14% and 18.4% of all deliveries at medical centers and regional hospitals, respectively, and 0% at clinics. The high incidence at teaching hospitals could be reflecting clinical care provision by inexperienced residents (mean physician age is 2 years lower in medical centers) or could represent the sequelae of conservative clinical policies favoring vaginal delivery. Were these conservative policies justified? What are the trade-offs between short-term and long-term morbidity of cesarean delivery as opposed to birth canal injuries sustained during vaginal delivery? To date, there is no conclusive data on the long-term impact on women’s physical and psychological health and life-satisfaction (including sexual functioning) following cesarean compared with vaginal delivery. Until the findings of well designed research become available, "norms" of cesarean delivery rates will remain controversial.
|
|
|
|
Footnotes |
|---|
doi:10.1097/01.AOG.0000102935.91389.53
Received June 11, 2003. Received in revised form September 16, 2003. Accepted September 18, 2003.
|
REFERENCES |
|---|
|
TOP
ABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
2. Belizan JM, Althabe F, Barros FC, Alexander S. Rates and implications of caesarean sections in Latin America: ecological study. BMJ 1999;319:1397–400.
3. Hopkins K. Are Brazilian women really choosing to deliver by cesarean? Soc Sci Med 2000;51:725–40.[Medline]
4. Morrison J, MacKenzie IZ. Cesarean section on demand. Semin Perinatol 2003;27:20–33.[Medline]
5. Obstetrics: rising cesarean rate is cause for alarm, say nurse-midwives. Med Lett CDC FDA 2003:23.
6. Evans L. Italy has Europe’s highest caesarean section rate. BMJ 1995;310:487.
7. 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.[Abstract]
8. World Health Organization.Appropriate technology for birth. Lancet 1985;2:436–7.[Medline]
9. Shearer EL. Cesarean section: medical benefits and costs. Soc Sci Med 1993;37:1223–31.[Medline]
10. Barros FC, Vaughan JP, Victora CG, Huttly SR. Epidemic of caesarean sections in Brazil. Lancet 1991;338:167–9.[Medline]
11. Gould JB, Davey B, Stafford RS. Socioeconomic differences in rates of cesarean section. N Engl J Med 1989;321: 233–9.[Abstract]
12. Placek PJ, Taffel SM. Recent patterns in cesarean delivery in the United States. Obstet Gynecol Clin North Am 1988;15:607–27.[Medline]
13. Haas JS, Udvarhelyi S, Epstein AM. The effect of health coverage for uninsured pregnant women on maternal health and the use of cesarean section. JAMA 1993;270:61–4.[Abstract]
14. Hueston WJ, Rudy M. A comparison of labor and delivery management between nurse midwives and family physicians. J Fam Pract 1993;37:449–54.[Medline]
15. Burns LR, Geller SE, Wholey DR. The effect of physician factors on the cesarean section decision. Med Care 1995; 33:365–82.[Medline]
16. Evans MI, Richardson DA, Sholl JS, Johnson BA. Cesarean section. Assessment of the convenience factor. J Reprod Med 1984;29:670–6.[Medline]
17. Localio AR, Lawthers AG, Bengtson JM Hebert LE, Weaver SL, Brennan TA, et al. Relationship between malpractice claims and cesarean delivery. JAMA 1993; 269:366–73.[Abstract]
18. Oleske DM, Glandon GL, Giacomelli GJ, Hohmann SF. The cesarean birth rate: influence of hospital teaching status. Health Serv Res 1991;26:325–37.[Medline]
19. McKenzie L, Stephenson PA. Variation in cesarean section rates among hospitals in Washington State. Am J Public Health 1993;83:1109–12.
20. Anderson GM, Lomas J. Determinants of the increasing cesarean birth rate. Ontario data 1979 to 1982. N Engl J Med 1984;311:887–92.[Abstract]
21. Stafford RS. Cesarean section use and source of payment: an analysis of California hospital discharge abstracts. Am J Public Health 1990;80:313–5.
22. Taffel SM, Placek PJ, Liss T. Trends in the United States cesarean section rate and reasons for the 1980–85 rise. Am J Public Health 1987;77:955–9.
23. Althaus F. Previous cesarean delivery and abnormal labor are main factors in high US cesarean section rate. Fam Plann Perspect 1994;26:186–7.
24. Turner RM. Cesarean section rates: reasons for operations vary between countries. Fam Plann Perspect 1990;22:281–2.
25. Di Lallo D, Perucci CA, Bertollini R, Mallone S. Cesarean section rates by type of maternity unit and level of obstetric care: an area-based study in central Italy. Prev Med 1996; 25:178–85.[Medline]
26. Lin YY. The impact of National Health Insurance on physician productivity and income (Chinese). Hospital 1999;32:1–14.
27. Tsai YW, Hu TW. National health insurance, physician financial incentives, and primary cesarean deliveries in Taiwan. Am J Public Health 2002;92:1514–7.
28. Potter JF, Berquo E, Perpetuo IH, Leal OF, Hopkins K, Souza MR, et al. Unwanted cesarean sections in Brazil: prospective study. BMJ 2001;323:1155–8.
This article has been cited by other articles:
|
|
 |
|
  S. Xirasagar, H.-C. Lin, and T.-C. Liu Do group practices have lower caesarean rates than solo practice obstetric clinics? Evidence from Taiwan. Health Policy Plan., July 1, 2006; 21(4): 319 - 325. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H.-C. Lin, S. Xirasagar, and Y.-C. Tung Impact of a cultural belief about ghost month on delivery mode in Taiwan. J Epidemiol Community Health, June 1, 2006; 60(6): 522 - 526. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
