Management and Outcomes for Children with Pyloric Stenosis Stratified by Hospital Type

Results The mean total preoperative LOS was twice as long at our public hospital than our private hospital (1.78 d versus 0.85 d, P = 0.003). Similarly, the mean total length of stay was 0.87 d longer at our public hospital (3.87 d versus 3 d), which trended toward significance ( P = 0.06). Significantly, 72% (27/33) patients at the private hospital had their US either as an out-patient or on the day of admission, while only 58% (28/48) had an US performed as expeditiously at the public hospital ( P = 0.03, χ 2 analysis). Conclusions Care at our public hospital leads to delayed diagnosis and treatment, resulting in longer preoperative and total LOS in infants with pyloric stenosis, most likely related to the timing of the preoperative US. Efforts should be made to encourage patients to use the outpatient clinic facilities for their ill infants, or for physicians at public hospitals to expedite the US process once the diagnosis of pyloric stenosis is entertained. Key Words health care disparities infants pyloric stenosis Introduction With the populations of uninsured and underinsured Americans at 45.7 million and 25.2 million, respectively, access to healthcare is one of the critical issues facing the American medical community today [1, 2] . In the pediatric population, a survey from 2007 revealed that 11% of children (8.1 million) were without health insurance, and children in poverty were significantly more likely to be uninsured [1] . The role of barriers to health care access has been evaluated in a variety of diseases, and for children the disease model used most commonly is appendicitis [3–6] . Appendicitis has no known risk factors and the incidence is fairly uniform across race, gender, ethnicity, and socioeconomic status [5] . However, while appendicitis is a frequent disease among school age children, it rarely occurs in infants. Furthermore, the surgery for appendicitis in children is often performed by an adult general surgeon rather than a pediatric surgeon, which may magnify the effect of health care disparities, since access to specialty care is more complex [7] . Therefore, the impact of health care disparities on infants, perhaps the most vulnerable subpopulation of children, has not been well characterized. Thus, a study of a disease that primarily afflicts infants will allow better evaluation of whether disparities exist for our youngest patients. Pyloric stenosis is an excellent model for addressing the impact of, access to, quality of, and outcomes for healthcare between different demographic groups, since it is a common diagnosis among infants. It affects 1.5 to 4 out of 1000 infants in the United States, with no known risk factors, although some data suggest it is more common among first-born males [8] . The treatment for pyloric stenosis is a pyloromyotomy following appropriate fluid and electrolyte repletion. The surgery is curative with complete recovery and discharge expected within 48 h of the procedure. While the outcomes for this disease have been previously documented, there are no studies directly contrasting outcomes among different socioeconomic groups where the surgeons and other health care providers managing the patients have been the same. Therefore, we reviewed our experience of all infants diagnosed with pyloric stenosis at both our urban public and university teaching hospitals. Our hypothesis was that while access to health care may be delayed at our public hospital, necessitating a longer preoperative stay, the overall health outcomes once the patient is in the healthcare system would be uniform. Methods All children who have undergone pyloromyotomy at our urban public and university teaching hospitals from July 31, 1998 to July 31, 2008 were included in the initial analysis. Only those with the confirmed intraoperative diagnosis of pyloric stenosis and with a complete and available medical record were included in the final analysis. Data abstracted included gestational age at delivery, birth weight, gender, race, insurance status, age at presentation, symptom duration, preoperative ultrasound (US), operative date, laboratory evaluations, as well as preoperative, postoperative, and total length of stay (LOS). Patients were initially divided into groups based on their hospital type, and then by insurance status and age at presentation to evaluate outcome variables. Surgical and medical care at the two institutions is rendered by the same pediatric and surgical residents as well as the same pediatric radiologists and radiology residents. The pediatric surgical residents cover both hospitals while on rotation, however, the patients at the private hospital are admitted to the pediatric surgery service while patients at the public hospital are admitted to the pediatrics service. Twenty-four-hour access to US is available at both institutions by a single radiology resident on-call at night and on weekends, but no pediatric radiologist is available at these times. Data represent mean ± standard deviation for continuous variables. Categorical data was analyzed using χ 2 analysis or Fisher's exact test, if the expected value of any cell was less than five. Continuous variables were compared using the Student's t -test when the data were normally distributed, or the Wilcoxon Rank Sums test if the data were skewed. Statistical significance was assigned to P values ≤ to 0.05. Results There were 110 patients with complete medical records and the diagnosis of pyloric stenosis during the study period, 55 at each hospital. The patient demographics at the private and public hospitals were vastly different. Nearly 73% of children at the private hospital were Caucasian compared with 0% at the public hospital ( P < 0.001), and nearly 75% of children at the private hospital had private insurance compared with 11% at the public hospital ( Table 1 ). Children at the public hospital had an older gestational age in wk (39.7 ± 1.64 versus 38.1 ± 1.9, P = 0.001) and higher birth weight in g (3748 ± 1006 versus 3004 ± 621, P = 0.01) compared with those at the private hospital. However, the age at presentation in days for these patients tended to be younger (30.3 ± 16.8 versus 35.7 ± 15.4, P = 0.08) rendering the postconceptual age at presentation nearly equivalent in the two groups. Similarly, there was no difference in gender distribution or duration of symptoms between patients at the public and private hospitals ( Table 1 ). Additionally, there was no statistical difference in all but one laboratory value at the time of presentation despite the evaluations being performed at two separate laboratories ( Table 2 ). The only statistical difference was a slightly higher serum chloride value at the public hospital (103 ± 5.6 versus 100 ± 6.4, P = 0.005), the clinical impact of which is unknown. Interestingly, the mean total preoperative LOS was twice as long at our public hospital than at our private hospital (1.78 ± 2.1 versus 0.85 ± 0.62, P = 0.003). There was no difference in postoperative LOS between the two groups (2.08 ± 1.3 versus 2.15 ± 0.85, P = 0.74). The mean total LOS was 0.87 d longer for infants cared for at our public hospital than infants at the private hospital, which trended toward but did not reach statistical significance (3.87 ± 3.1 versus 3.00 ± 1.0, P = 0.06). These results are detailed in Table 3 . In an effort to explain the increased preoperative LOS at the public hospital, we evaluated the timing of the patients' presentation, US, and operation. Significantly, 72% (27/33) of the patients at the private hospital had their US either as an out-patient or on the day of admission while only 58% (28/48) had an US performed as expeditiously at the public hospital ( P = 0.03, χ 2 analysis). There was no difference in the time from US until operation between the two groups (0.958 ± 1.35 versus 0.818 ± 0.727, P = 0.55). Stratification by age at presentation and insurance status made no difference in the outcome variables assessed. Discussion Healthcare disparities in underserved communities remain an important public health topic. Infants represent perhaps a most vulnerable population since they continue to develop well into the postnatal period, and there are a number of medical conditions that afflict infants during this crucial time. A delay in healthcare may be detrimental to babies who suffer from such conditions. Barriers to healthcare for children in general have been identified; yet, the impact of healthcare disparities on infants in particular has not been well characterized. Since pyloric stenosis is a common condition in infants with no known risk factors, it provides a useful model for evaluating health care disparities in this age group. Our Division of Pediatric Surgery serves two distinct populations at the private and public city hospitals and, thus, it offers a unique opportunity to study health disparities in two distinct patient populations while standardizing diagnosis and management algorithms. Gupta et al. concluded that poor children, who are either uninsured or have Medicaid, have more frequent and longer hospitalizations and, as a result, higher costs than non-poor children [9] . The authors proposed that differences in severity of disease for indigent and minority children on admission are a possible etiology for the disparities. Similarly, it has been reported that appendicitis patients with barriers to healthcare access delay seeking care, ultimately presenting with a greater incidence of perforation and associated complications [10] . We also found that children with appendicitis at our public city hospital have longer lengths of stay and undergo more expensive imaging [11] . However, our retrospective review of infants with pyloric stenosis at two urban hospitals demonstrated that there was no significant clinical difference in the infants at presentation, suggesting that the parents of both groups sought care at similar stages of the disease. The patients at both hospitals experienced symptoms for the same duration, and laboratory values at presentation showed no statistical significance, with the exception of chloride, which was slightly higher in the public hospital. The significance of this finding is unclear. However, the lack of any disparity in serum bicarbonate values indicates that the patients at the public hospital did not require more rehydration prior to surgery that could have possibly resulted in their longer preoperative LOS. Despite the similarities at presentation, the preoperative LOS was greater at the public hospital compared with the private hospital, with a trend toward an increased total LOS, and no difference in the postoperative LOS. This increase in preoperative LOS could be due to either delays in diagnosis or delays in surgical management. In general, diagnosis of pyloric stenosis is determined primarily by patient history, physical exam, and increasingly radiographic imaging. Ward et al. noted that the absence of metabolic derangements and, therefore, of laboratory abnormalities can make the diagnosis of pyloric stenosis more difficult [8] . However, we did not find any statistically significant difference in metabolic derangements between patients seeking care at our two urban hospitals. In evaluating timing of patient admission, US, and operative intervention, we determined that the greater preoperative LOS at the public hospital results from a significantly higher proportion of infants who underwent US at least one night after their admission to the hospital. Common practice at our public hospital is that babies are admitted from the pediatric emergency room (ER) without a definitive diagnosis, and then receive an US as an inpatient, which may not occur until the next day. On the contrary, infants that are admitted to our private hospital usually receive an US as an outpatient prior to admission or immediately upon admission to our hospital. This seems to be the major mechanism behind the difference in LOS between the two hospitals, and strategies to limit use of the ER as a primary care option may abrogate this effect. An alternative strategy would be to increase access to a pediatric radiologist, since unavailability at night is the most common reason that the US is delayed until the next morning, although the cost of implementing teleradiology or of increasing in-house hours for the radiologists may outweigh the cost of the overnight admission. Medicaid beneficiaries make almost 20% of all ER visits and, in particular, Medicaid children are noted to have rising rates of ER use [12] . Fredrickson et al. found that ER use was concentrated in a subgroup of children with half of all visits made by 17% of the Medicaid-insured children who used any emergency or hospital services for asthma [13] . This increased use of the ER is closely tied to reduced access to a regular source of care, especially primary care providers. Besides difficulties in obtaining urgent advice or appointments, parents frequently had difficulties in transport, childcare, or work release that impeded using the primary care office as the principal site of asthma care [13] . This mechanism could certainly be at play in our study, since less than 60% of patients at our city hospital received an US on the day of admission. Our analysis did not specifically assess this, but lack of access to primary care and resultant use of the ER could explain the extended preoperative LOS at our public hospital. A potential limitation of our analysis is that the preoperative LOS data for the public hospital included time spent at outside hospitals prior to being transferred to our city hospital. There are inevitable delays in care for patients who require transfer [12] . In contrast, the preoperative LOS data for our private hospital represents patients who came either from home or the pediatrician's office, and nearly three-quarters of those who had an US had it on the day of admission. However, all patients who were transferred from outside hospitals required either a repeat US or an official reading from our pediatric radiologists; therefore, the time to US remained a factor contributing to delays in definitive care in all patients. In summary, our results provide evidence that delays in definitive diagnosis in our public hospital system and not late presentation of infants with pyloric stenosis may be responsible for an increased LOS. While we did not evaluate hospital charges or costs, certainly an increase in LOS will result in greater expense to the healthcare system. Outpatient diagnosis should enable prompt treatment, with the expectation of shortened inpatient LOS. Efforts should be made to encourage patients to use the outpatient clinic facilities or for physicians in the ER at public and private hospitals alike to expedite the US process. Once accurate and definitive diagnosis has been made, expeditious treatment can be provided to ensure the highest quality of care for infants, with the optimal utilization of resources. References 1 C. DeNavas-Waly B.D. Proctor J.C. Smith Income, Poverty, and Health Insurance Coverage in the United Sates: 2007 U.S. Census Bureau, Current Population Reports 60 2008 235 2 C. Schoen S.R. Collins J.L. Kriss How many are underinsured? Trends among U.S. adults, 2003 and 2007 Health Affairs Web Exclusive 298 2008 309 3 N.R. Burrows Y. Li D.E. Williams Racial and ethnic differences in trends of end-stage renal disease: United States, 1995 to 2005 Adv Chronic Kidney Dis 147 2008 152 4 R. Correa-de-Araujo B. Stevens E. Moy Gender differences across racial and ethnic groups in the quality of care for acute myocardial infarction and heart failure associated with comorbidities Women Health Issues 44 2006 55 5 K.A. Jablonski M.F. Guagliardo Pediatric appendicitis rupture rate: A national indicator of disparities in healthcare access Popul Health Metr 2005 4 6 D. Merenstein B. Egleston M. Diener-West Lengths of stay and costs associated with children's hospitals Pediatrics 839 2005 844 7 A.C. Skinner M.L. Mayer Effects of insurance status on children's access to specialty care: A systematic review of the literature BMC Health Serv Res 2007 194 8 E. Ward D. Easley J. Pohl Previously unsuspected infantile hypertrophic pyloric stenosis diagnosed by endoscopy Dig Dis Sci 946 2008 948 9 R.S. Gupta M. Bewtra L.A. Prosser Predictors of hospital charges for children admitted with asthma Ambulatory Pediatr 15 2006 20 10 S. Eldar E. Nash E. Sabo Delay of surgery in acute appendicitis Am J Surg 194 1997 198 11 N.R. Zilbert E.F. Stamell I. Ezon Management and outcomes for children with acute appendicitis differ by hospital type: Areas for improvement at public hospitals Clin Pediatr (PA) 48 2009 499 12 R.W. Bryne B.T. Bagan K.V. Slavin Neurosurgical emergency transfers to academic centers in Cook County Neurosurgery 709 2008 716 13 D.D. Fredrickson C.A. Molgaard S.A. Dismuke Understanding frequent emergency room use by Medicaid-insured children with asthma J Am Board Fam Pract 96 2004 100