Cesarean section rates and outcomes relative to hospital size

¶ … C-Sections Relative to Hospital Size

A cesarean section, more commonly known as a C-section, is a surgical procedure in which a fetus is delivered via a hysterotomy rather than a more typical vaginal delivery procedure. There are many medical reasons why a c-section is necessary, although the geographic and demographic distribution of c-sections shows many trends in the prevalence of the procedure, indicating medical necessity is not the only reason that the surgery is performed. Affluence and medical availability and infrastructure have clear impacts on the rate and outcome of c-sections.

In 2010, the World Health Organization estimated that on average, 10-15% of live births were via c-section, and the rate of c-sections in the United States, despite a rising popularity of alternative birthing methods (midwifes, home births, doulas, etc.), the c-sections have increased by 48% since 1996. While the size of hospitals studied when studying c-section outcomes has been noted, analysis is lacking on the data collected to determine c-section outcomes relative to the size of the hospital.

What was the hypothesis?

The hypothesis in the study by Kritchevsky et al., published in the American Journal of Obstetrics and Gynocology was that due to lack of standardized definitions and performance measurement systems, comparative study of c-sections across different hospitals was not fully possible. The study by Lagrew et al., was designed to examine delivery statistics obtained from a cesarean section reduction program conducted in a private hospital. The authors hypothesized that this program had been successful in reducing the number of c-sections within the hospital.

A strong study (Chen et al. 2008) using a large sample number and thorough quantitative analysis could serve as a useful model for studying the impact of urbanization and population density upon the rate of cesarean deliveries. Are c-section deliveries, which are faster, seen more often in urban hospitals with a large demand for birthing space? A large-scale study in Taiwan published in 2007 examines association.

The authors' hypothesized that there was a positive association between the degree of urbanization in Taiwan and the likelihood of a Cesarean section for the delivery of a single fetus.

What study design was implemented (cohort, case control, cross-sectional etc.)?

The study by Kritchevsky et al. (1996) looked at outcomes based on retrospective data abstraction of 200 deliveries at each of 15 acute care hospitals, including two foreign institutions. The second study (Lawgrew et al.) employed retrospective analysis of data obtained from a c-section reduction program conducted within the same institution.

The third study was a retroactive analysis of 200,207 singleton deliveries in Taiwan using information from the Taiwan 2004 National Health Insurance Research Database. The urbanization level of the mother's home area was classified into 7 categories. The data was examined with a multilevel logistic regression model to show the assocation of urbanization with rate of c-section (Chen 2008).

How were the variables operationalized?

Figure 1 in the study (Kritchevsky et al. 1996) demonstrates the way in which the variables were operationalized by normalizing data using various instruments into a relatively comparable measure. The process of standardizing the measures assisted in defining the variables further by indicating the types and degree of difference between individual institution's measures. The variables in the Lagrew et al. (1996) study were easily operationalized due to the fact that this was a study that was done retrospectively to determine if the implemented changes had the desired effect on the rates of c-section deliveries in the hospital.

The variables in the third study (Chen et al. 2008) were operationalized by deciding upon a 7 point scale of urbanization. The creation of categorical variables allowed for results ordered in an easily-understood gradient rather than using a more abstract instrument for measuring urbanization. The operationalization of c-section rates was in accordance with standard definitions used in published literature.

Were the instruments validated (i.e. against measures obtained using a different instrument)?

The retrospective data analysis methodology made validation in the study published by Krichevsky et al. 1996 impossible, as many different measures were used by the hospitals in the process of their own data collection. This article looking at hospital performance and Cesarean section rates looked explicitly at hospital sizes, among other criteria, focused predominantly on medium and large-sized medical facilities -- with an average of 537 beds -- in the United States. The size of the hospital can have some indication on the level of care provided at the facility, but smaller facilities specializing in neonatal care exist. This study was looking at facilities offering comprehensive medical services, but within the domain of tertiary medical services, more than half of the facilities studied had neonatal intensive care units.

One of the complicating factors in comparing c-section outcomes between hospitals stemmed from differences in birth outcome coding procedures between hospitals, despite standard procedures established by the U.S. Department of Health and Human Services (Kritchevsky 2008 288). The authors noted that there is "no national standardized specifications for calculating a Cesarean section rate (Kritchevsky 288)." The variations between hospitals in their indicators and coding systems resulted in the authors advocating for standardization in order to facilitate comparative judgments for which there is "widespread demand for external release of outcome data from hospitals to insurers, employers, legislators, consumer advocates, regulatory agencies, accrediting bodies and many others (Kritchevsky 290)."

In the second study reviewed that was published in the same journal in 1996, the instruments were not validated, perhaps because the study (Lagrew et al. 1996) was designed to look at the potential for non-mandated clinical changes impacting the c-section rate in one particular private hospital. This was not a comparative study with another institution or population. The hypothesis was, in the authors' eyes, proven by the statistical analysis of the data using a solitary measurement. In the study done by Chen et al. In 2008 on Taiwanese births, a singular instrument was used in this study, the multilevel logistic regression model. The regression model was adjusted for demographic factors that could have had an impact on the data by presenting outliers, however