See “Direct Costs of Acute Recurrent and Chronic Pancreatitis in Children in the INSPPIRE Registry” by Uc et al on page 443 and “Recurrent Pancreatitis in Children—More Data Still Needed” by Bennett on page 362.
What Is Known
- Acute pancreatitis in children is a problem of increasing magnitude.
- Acute pancreatitis is recognized to be a condition associated with a low mortality in children.
- Acute recurrent pancreatitis occurs in 15% to 35% of children following an initial episode of acute pancreatitis.
What Is New
- The overall mortality for children with acute pancreatitis was 0.4%, which was significantly less than that determined by previous pediatric studies (5%–11%).
- A significant proportion (42%) of index cases of acute pancreatitis subsequently experienced an episode of acute recurrent pancreatitis with a median of 2 episodes.
- Patients with acute recurrent pancreatitis were younger and experienced a more severe index episode of acute pancreatitis.
Acute pancreatitis (AP) in children is a problem of increasing magnitude (1–3). It is a disease with a considerable health care burden in hospitalized (4) and outpatient settings (5). The natural course of the disease in children remains unclear, specifically whether an index episode of AP predisposes children to future recurrent episodes of the disease.
Acute recurrent pancreatitis (ARP) is defined as ≥2 distinct episodes of AP with an intervening return to baseline (6). Limited data from single-center studies indicate that ARP occurs in 15% to 35% of children following an initial episode of AP (7). An accurate determination of ARP incidence in children bears important implications for the evaluation of index episodes of AP. We therefore interrogated a large national database with the primary aim of determining the incidence of ARP in children.
We used data from a multicenter, administrative, national dataset, the Pediatric Health Information System (PHIS). The PHIS database is powered by 44 not-for-profit, free-standing, tertiary care children's hospitals in the United States. Discharge data from these participating hospitals include demographic information and diagnoses coded with the International Classification of Diseases, Ninth Revision (ICD-9) codes.
Patients from 1 to 17 years of age at a PHIS-participating hospital from January 2002 through December 2014 were included in our study. We submitted our study protocol to The University of Missouri institutional review board. It was determined exempt from the institutional review board review process.
For the purposes of the present study, we used the following definitions:
- Single episode of AP: Patients were recorded in the database with only a single AP-related admission for the duration of the study.
- ARP: This entailed a repeat AP-related admission with an intervening period of >14 days between any 2 AP-related hospitalizations.
- Index hospitalizations for AP: These represented the initial AP-related admission for any single patient. Subsequently, these patients may have experienced a repeat AP-related admission (these were termed ARP) or not (these were termed as a single episode of AP).
Data collection included demographic information including age, sex, and length of hospital stay (LOS); hospital charges rounded to the nearest $1000; mortality; and admission to the intensive care unit (ICU). The discharge diagnosis of AP was identified using ICD-9 code 577.0, which is the sole code for this condition. For the purpose of the present study, we extracted all cases with any (primary or secondary) discharge diagnosis of AP. This methodology was agreed on among the authors to prevent the exclusion of patients with AP, which may have received alternate primary-level diagnoses such as abdominal pain, nausea, vomiting, and so on. Race was not included because these data were missing for a significant number of cases.
We adapted the definition of ARP for use with an administrative database to ≥2 distinct episodes of AP with an intervening period of >14 days between hospitalizations. Any readmits that occurred within 14 days of a prior AP-related hospitalization were excluded from the present study.
Continuous variables were summarized as median and interquartile range (IQR), whereas categorical variables were presented as numbers and percentages. Categorical variables were compared using a chi-squar test of association, whereas continuous variables were compared using a Wilcoxon rank-sum test. Statistical analyses were performed using SAS version 9.3 (SAS Institute, Cary, NC).
From 2002 to 2014, there were a total of 5,060,555 hospital discharges in children 1 to 17 years of age, of which 26,435 children had a diagnosis of AP (Table 1). The median age of children with AP was 12 years (IQR 8–15 years). The median LOS was 4 days (IQR 2–7 days) and the median medical hospital charges amounted to $20,000 (IQR $11,000–$42,000). Mortality in this cohort was 0.4%, whereas ICU stay was recorded to be 9.0%.
In the cohort of patients with AP, 10,648 discharges were index hospitalizations for AP. Of these, 6159 children were recorded with a single hospitalization for AP, whereas 4489 children underwent a total of 15,787 rehospitalizations. These rehospitalizations were deemed as episodes of ARP because they all occurred >14 days after the previous AP-related admission. The cohort of 4489 children underwent a median of 2 (IQR 1–4) ARP-related hospitalizations with a range from 1 to 52 episodes. Median time between ARP-related admissions was 86 days (IQR 37–218 days) with a range from 15 to 3305 days (Table 2).
We compared the characteristics of patients with a single episode of AP to patients with subsequent ARP-related episodes, as assessed at the time of their index hospitalization. Compared with their counterparts with only a single episode of AP, patients with ARP were younger (median age 11 years [IQR 7–14 years] vs 13 years [IQR 9–15 years]; P < 0.01), had a longer LOS (median stay 5 days [IQR 3–10 days] vs 4 days [IQR 2–6 days]; P < 0.01), and had higher hospital charges (median charges $26,000 [IQR $13,000–$58,000] vs $17,000 [IQR $10,000–$34,000]; P < 0.01). Although there were no significant sex differences between the 2 groups (P = 0.20), children with ARP demonstrated an increased requirement for ICU care (risk ratio [RR] 2.1 [1.8–2.4]; P < 0.01).
We also compared AP-related index hospitalizations with ARP-related hospitalizations (all subsequent visits). Compared with hospitalizations for AP, those related to ARP were slightly shorter (median 3 days [IQR 2–6 days] vs median 4 days [IQR 2–8 days]; P < 0.01) but equal in cost (median $21,000 [IQR $11,000–$41,000] vs $20,000 [IQR $11,000–$43,000); P = 0.53). Also, children with an ARP-related hospitalization had an increased requirement for ICU care compared with children with an index episode of AP, RR 1.3 (1.2–1.5); P < 0.01. No significant difference in mortality was, however, noted between ARP- and index AP-related hospitalizations (0.50% vs 0.30%; RR 1.3 [0.9–2.0]; P = 0.18).
To our knowledge, the present study is the first study of its kind that has used multiinstitutional data to specifically determine the incidence of ARP in children. The strengths of the study include a large cohort of patients from 44 children's hospitals, greatly reducing selection and regional biases.
Although AP is recognized to be a condition associated with a low mortality in children (8), we observed that the overall mortality was 0.4%, which is significantly less than that determined by the previous pediatric studies (5%–11%) (3,9,10). It is unclear why this discrepancy exists; however, it could be related to the inclusion of older data from the 1990s by other studies. It should be emphasized here that deaths in children are rarely attributable to AP, but instead because of underlying comorbidities and diseases. Our results also demonstrate that 1 of every 10 children with AP experienced ICU care. We could not determine how many of these patients required ICU care as a consequence of the severity of their disease as opposed to another concurrent medical condition.
The median LOS in our study was 4 days, which was somewhat shorter than the median of 5 to 8 days reported by other studies (9,11,12). Again, this difference could be explained by regional biases and inclusion of older data by previous studies.
A significant proportion (42%) of index cases of AP subsequently experienced an episode of ARP. There was a median of 2 episodes of ARP in these patients, with the first episode occurring in a relatively short period (median of 86 days). The higher recurrence is at odds with previous studies having indicated a much lower occurrence of ARP in children. Benifla and Weizman (10) reviewed a total of 589 pediatric patients with AP in the literature and determined that 9% experienced a recurrence. Other US studies have reported a recurrence of 12% (9) and 15% (2). In contrast, a Mexican study observed a much higher recurrence of 34.5% (13). Also, an Italian study has reported a recurrence of 28% (14).
We attempted to identify factors that would predict those individuals who experienced ARP. Our results suggest that these patients are younger, and experienced a more severe index episode of AP as evidenced by longer LOS, higher hospital charges, and greater proportion of patients requiring ICU stay. Although these differences are interesting, in the absence of regression modeling these data should not be used for risk stratification of patients. ARP-related rehospitalizations were generally similar to index episodes of AP except that they were slightly shorter in LOS and had a somewhat higher requirement for ICU care.
Our results merit reconsideration of the generally accepted practice in pediatric AP wherein an index episode of pediatric pancreatitis is rarely investigated beyond the use of routine laboratory tests and ultrasound imaging. The results of the present study, however, demonstrate that almost half of the children with an index episode of AP will present with a repeat episode. Furthermore, children with a single ARP-related hospitalization were likely to experience further recurrent episodes. These data suggest a role for a greater degree of testing in children presenting with an ARP-related admission. This may involve magnetic resonance imaging or computed tomography of the pancreas to evaluate for anatomical anomalies and testing for genetic mutations known to predispose children to ARP (15).
The chief limitations of the present study relate to its retrospective nature and the sole reliance on ICD-9 codes for case identification. ARP-related admissions to non-PHIS hospitals were unknown but could have had a significant impact on the results of this study. Furthermore, we did not investigate the etiology of AP. Finally, we did not perform multivariate regression analyses for outcomes such as mortality, LOS, and hospital charges.
AP in children is associated with a high incidence of ARP. Hospitalizations related to recurrent episodes of AP occurred with greater frequency in children with a severe index episode of AP. ARP-related hospitalizations in children were generally comparable in outcomes to index episodes of AP and overall mortality remained very low. Further investigation is needed to determine whether an initial presentation of AP in children merits further etiological investigation.
1. Morinville VD, Barmada MM, Lowe ME. Increasing incidence of acute pancreatitis
at an American pediatric tertiary care center: is greater awareness among physicians responsible? Pancreas
2. Park A, Latif SU, Shah AU, et al. Changing referral trends of acute pancreatitis
: a 12-year single-center analysis. J Pediatr Gastroenterol Nutr
3. Nydegger A, Heine RG, Ranuh R, et al. Changing incidence of acute pancreatitis
: 10-year experience at the Royal Children
's Hospital, Melbourne. J Gastroenterol Hepatol
4. Pant C, Deshpande A, Olyaee M, et al. Epidemiology of acute pancreatitis
in hospitalized children
in the United States from 2000–2009. PLoS One
5. Pant C, Deshpande A, Sferra TJ, et al. Emergency department visits for acute pancreatitis
: results from the nationwide emergency department sample 2006–2011. J Investig Med
6. Morinville VD, Husain SZ, Bai H, et al. Definitions of pediatric pancreatitis and survey of present clinical practices. J Pediatr Gastroenterol Nutr
7. Meyer A, Coffey MJ, Oliver MR, et al. Contrasts and comparisons between childhood and adult onset acute pancreatitis
8. Goday PS, Werlin SL, Wakeham M, et al. Acute pancreatitis
in the pediatric intensive care unit. J Pediatr Gastroenterol Nutr
9. Werlin SL, Kugathasan S, Frautschy BC. Pancreatitis in children
. J Pediatr Gastroenterol Nutr
10. Benifla M, Weizman Z. Acute pancreatitis
in childhood: analysis of literature data. J Clin Gastroenterol
11. Kandula L, Lowe ME. Etiology and outcome of acute pancreatitis
in infants and toddlers. J Pediatr
12. Park AJ, Latif SU, Ahmad MU, et al. A comparison of presentation and management trends in acute pancreatitis
between infants/toddlers and older children
. J Pediatr Gastroenterol Nutr
13. Sanchez-Ramirez CA, Larrosa-Haro A, Flores-Martinez S, et al. Acute and recurrent pancreatitis in children
: etiological factors. Acta Paediatr
14. Pezzilli R, Morselli-Labate AM, Castellano E, et al. Acute pancreatitis
. An Italian multicentre study. Dig Liver Dis
15. Minen F, De Cunto A, Martelossi S, et al. Acute and recurrent pancreatitis in children
: exploring etiological factors. Scand J Gastroenterol