Ever since the seminal study by Luft examining the inverse association between hospital volume and outcomes that was published about 35 years ago, numerous reports have unequivocally reinforced that high hospital volumes are associated with better outcomes for a wide range of surgical procedures (Luft, 1980; Mariotto, 2001; Birkmeyer et al., 2003; Dimick et al., 2005; Chang and Birkmeyer, 2006; Kahn et al., 2006; Balentine et al., 2014). This relationship is particularly pronounced for extremely complex surgical procedures, which are in low numbers overall and usually constitute only a fraction of the annual caseloads of hospitals (Studnicki et al., 2014). Several stakeholders including providers, policy makers, insurance companies, and patient advocates have suggested that highly complex procedures be selectively referred or regionalized to select centers of excellence to realize the best possible outcomes (Leapfrog Group, 2006). These select centers of excellence tend to be high-volume, highly specialized hospitals and teaching hospitals. Regionalization is the increase in concentration of procedures into select centers. Over the course of the past decade, this phenomenon of regionalization has been demonstrated across multiple surgical specialties and procedures (Ben-David et al., 2012; Pinto et al., 2012; Simhan et al., 2012; Colavita et al., 2013; Hayman et al., 2013; Smaldone et al., 2013; Colavita et al., 2014).

Craniosynostosis is a congenital defect wherein there is premature fusion of cranial sutures and a resultant alteration of the growth pattern of the skull and increased intracranial pressure (Slater et al., 2008). Estimates of craniosynostosis range from 3 to 13 per 10,000 live births (French et al., 1990; Boulet et al., 2008). Typically, early surgical intervention is advocated to minimize the risks of complications associated with increased intracranial pressure (Panchal and Uttchin, 2003). This is a complicated surgical procedure, with only select hospitals performing it. To date, there is no empirical evidence on whether surgical repairs for craniosynostosis have followed this regionalization pattern.

The objective of the present study is to examine whether surgical repairs for craniosynostosis were regionalized to teaching hospitals over the 8-year period from 2003 to 2010. The present study tests the hypothesis that over the study period, more surgical procedures were performed in teaching hospitals as opposed to nonteaching hospitals.

Materials and Methods

Results

During the study period (years 2003 to 2010), a total of 19,417 patients aged up to 3 years underwent a surgical repair for craniosynostosis. The characteristics of these patients are summarized in Tables 1 and 2. The most frequently performed surgical procedure was cranial osteoplasty (61.5% of patients), followed by opening of cranial suture (36%), bone graft to skull (19.5%), and formation of cranial bone flap/repair of skull with flap (14.2%). A vast majority of patients underwent surgery before the age of 1 year (76.6% of all procedures). Of all patients, 65.3% were boys and 34.7 were girls. Whites were the predominating race (65.1%), followed by Hispanics (20.6%). Private insurance plans covered 54% of patients; whereas, Medicaid covered 39.8%. Following the surgical procedure, 98.1% were discharged routinely. A majority (74.1%) of the patients did not have a concomitant comorbid condition, 20.2% had one comorbid condition, 4.8% had two comorbid conditions, 0.7% had three comorbid conditions, and 0.2% had four or more comorbid conditions. The most prevalent comorbid conditions were deficiency anemias (10.5%), fluid and electrolyte disorders (8.6%), coagulopathy (4.1%), neurological disorders (3.4%), and chronic blood loss anemia (1.1%).

View larger version(19K)

TABLE 1Characteristics of Patients Who Had Surgical Repair for Craniosynostosis

View larger version(22K)

TABLE 2Comorbid Burden

Performance of procedures by teaching hospitals over the course of the study period are presented in Table 3. The number of surgical procedures increased during the study period. It ranged from 1628 procedures in the year 2003 to 3001 procedures during the year 2010. The most procedures were performed in the year procedures). Overall, 90.2% of all procedures were performed in teaching hospitals and 9.8% were performed in nonteaching hospitals. Over the course of the study period, more procedures tended to be concentrated in teaching hospitals. Data show that 83.3% of all procedures in the year 2003 were performed in teaching hospitals; whereas, 97.5% of all procedures in the year 2010 were performed in teaching hospitals. During the years 2003 to 2006, close to 16% of all procedures were performed in nonteaching hospitals. During the years 2007 to 2010, only about 3% of all procedures were performed in nonteaching hospitals.

View larger version(4K)

TABLE 3Teaching Status of Hospital by Year

Estimates from the multivariable logistic regression model examining the association between year of procedure and performance in a teaching hospital are summarized in Table 4. Following adjustment for the effects of age, sex, race, and comorbid burden, there was a significant association between the year 2010 and performance of a procedure in a teaching hospital. Year 2010 was associated with increased odds of having the surgical procedures performed in a teaching hospital as opposed to a nonteaching hospital when compared with year 2003 (odds ratio = 10.43, 95% confidence interval, 1.10 to 98.98; P = .04).

View larger version(19K)

TABLE 4Odds of Being Treated in a Teaching Hospital—Estimates From Multivariable Logistic Regression Model

Discussion

The present study used the largest all-payer, nationally representative dataset of all U.S. hospitalizations to test the hypothesis of whether over an 8-year period (2003 to 2010) there was regionalization of surgical repair of craniosynostosis procedures into teaching hospitals. Results suggest that 83.3% of procedures in the year 2003 were performed in teaching hospitals; whereas, 97.5% of procedures in the year 2010 were performed in teaching hospitals. Over the course of this 8-year period, there was a 14.2% increase in regionalization of procedures into teaching hospitals. The association of increasing regionalization of surgical repairs for craniosynostosis into teaching hospitals was statistically significant even after adjustment for a multitude of patient-related factors such as age, sex, race, and comorbid burden. The present study findings are consistent with prior reports on regionalization for a wide range of surgical procedures (Ben-David et al., 2012; Pinto et al., 2012; Simhan et al., 2012; Colavita et al., 2013; Hayman et al., 2013; Smaldone et al., 2013; Colavita et al., 2014).

There are multiple benefits associated with regionalization of complex surgical procedures into select centers of excellence. Salazar et al. (2014) conducted a risk-adjusted examination of hospital surgical outcomes among children by geographic areas and found that for infants, surgery outcomes were significantly better when surgical procedures were performed in high-volume/specialized centers in multiple geographic regions. Austin and Urbach (2013) estimated that regionalization of colorectal surgery, esophagectomy, and pancreaticoduodenectomy significantly reduced perioperative deaths for these procedures. Colavita et al. (2014) examined outcomes following hepato-pancreato-biliary cancer resections and showed that regionalizing procedures into select centers not only reduced operative mortality rates but also reduced the length of stay in hospitals and improved the odds of being discharged routinely. Sundaresan et al. (2013) studied the feasibility and impact of regionalization of lung cancer resections in Ontario, Canada, and showed that concentrating these procedures into select centers facilitated delivery of more comprehensive care due to the ease of obtaining multidisciplinary and interdisciplinary assessments and care. They also showed that due to better processes of care, regionalization of pneumonectomy surgical procedures to select centers significantly reduced postsurgical 30-day mortality rates.

Although it is evident that regionalization of complex surgical procedures has resulted in better surgical outcomes, one should not discount the drawbacks of regionalization. Some of the issues associated with regionalization included reduced access to care, increased travel distances for patients, overloading of select hospitals, and lack of continuity of care. Pinto et al. (2012) examined the impact of distance on outcomes after discharge following neonatal congenital heart surgery and found that those living longer distances from the hospital that performed the surgical procedures were associated with worse morbidity and higher adverse events. Livingston and Burchell (2010) examined the effect on travel distance for Medicare patients who underwent bariatric procedures and found that travel time for patients increased by 76% following regionalization of procedures to select centers of excellence and may have reduced access to follow-up care. Furthermore, regionalization of procedures to only select centers may make the low-volume centers financially viable, and the former will be forced to stop services for less complex surgical procedures (Birkmeyer et al., 2003). This would eventually reduce access to care for a greater proportion of the populace. The costs associated with regionalization could be substantial from the patient's point of view due to the travel time involved, days off taken from work (for parents who are caring for children), and lack of periodic follow-up care after the surgical procedures. Finally, there is only a finite number of high-risk surgical procedures that a hospital can perform. Indiscriminate regionalization of all complex surgical procedures to select centers of excellence could make these centers extremely high volume, and their physical facilities may not be equipped to handle an overly high load of cases. They are likely to reach a point where they could experience diseconomies of scale that could compromise the process of care delivered to patients, which will eventually lead to poor outcomes.

The present study has multiple limitations attributed to the cross-sectional nature of the study design and use of large secondary discharge datasets such as the NIS to examine regionalization phenomenon. The significant findings demonstrated in the present study only indicate an association and not a true cause-and-effect relationship, which cannot be established with cross-sectional retrospective studies. The only way to do so would be to conduct prospective randomized controlled trials. However, this would be impractical, considering the ethical issues associated with randomization. Prior estimates suggest that regionalization of surgical procedures is a complex phenomenon, and a multitude of factors such as state-specific certificate-of-need laws, area-specific referral patterns, and market forces dictate the type of regionalization (formal versus informal) (Nallamothu et al., 2005; Glickman et al., 2010; Studnicki et al., 2014). The present study is unable to determine whether the observed regionalization is formal or informal due to the lack of availability of legal and market-specific variables in the NIS database. Risk adjustments conducted in the present study were limited to the variables available in the dataset, and consequently, we cannot rule out biases attributed to lack of data on a more comprehensive set of confounders. Despite the limitations, the present study findings are still valuable because the estimates are nationally representative, generalizable, and the first to examine the phenomenon of regionalization of surgical repairs of craniosynostosis on a national scale.

Conclusions

Results from the present study show that over the 8-year period from 2003 to 2010, an increasing proportion of surgical repairs of craniosynostosis were performed in teaching hospitals, suggesting an increasing concentration of these complex surgical procedures in select centers. The impact of regionalization of these procedures is unknown. As more longitudinal data become available, the relative benefits and drawbacks associated with regionalization of surgical repairs of craniosynostosis should be examined by future studies.