OBJECTIVE. Clostridium difficile is the main cause of nosocomial and antibiotic-associated diarrhea in adults. Recently, the incidence and severity of C difficile-associated disease in adults have been increasing. Whether similar phenomena are occurring among children remains unknown. Our study describes the epidemiological features of C difficile-associated disease in hospitalized children.
METHODS. We conducted a retrospective cohort study of hospitalized children with C difficile-associated disease at 22 freestanding children's hospitals in the United States, from 2001 to 2006. Cases of C difficile-associated disease were defined as a hospitalized child with a discharge code for C difficile infection, a laboratory billing charge for a C difficile toxin assay, and receipt of antimicrobial therapy for C difficile-associated disease.
RESULTS. We identified 4895 patients with C difficile-associated disease. Over the study period, the annual incidence of C difficile-associated disease increased from 2.6 to 4.0 cases per 1000 admissions and from 4.4 to 6.5 cases per 10 000 patient-days. The median age of children with C difficile-associated disease was 4 years. Twenty-six percent of patients were <1 year of age. The majority of patients (67%) had underlying chronic medical conditions. The colectomy and all-cause mortality rates among children with C difficile-associated disease did not increase during the study period.
CONCLUSIONS. The annual incidence of C difficile-associated disease in hospitalized children increased significantly from 2001 to 2006. However, the rates of colectomy and in-hospital death have not increased in children with C difficile-associated disease as they have among adults. The risk factors and outcomes for children with C difficile-associated disease remain to be defined in future studies.
Clostridium difficile is the most common cause of nosocomial and antibiotic drug-associated diarrhea in adults, with an estimated national cost of $1.1 billion per year.1,2 Reports have demonstrated increases in the incidence and severity of C difficile-associated disease (CDAD) in adults.3–5 CDAD varies from mild diarrhea to pseudomembranous colitis and toxic megacolon, which can cause significant morbidity and death. However, the epidemiological features of CDAD remain relatively undefined in the pediatric population. Studies suggest that C difficile is emerging as a major cause of childhood diarrhea in both community and hospital settings.5,6 Because there are no large epidemiological studies of CDAD in hospitalized children, defining the epidemiological characteristics of CDAD in children is of increasing importance, given the changes seen in adults.
We performed a retrospective cohort study, examining cases of CDAD among inpatients at 22 freestanding children's hospitals from 2001 to 2006.
We used the Pediatric Health Information System (PHIS), an administrative database that contains inpatient data from >40 not-for-profit, freestanding, tertiary care, children's hospitals in the United States. These hospitals are affiliated with Child Health Corporation of America (Shawnee Mission, KS). The PHIS contains detailed information on demographic characteristics, diagnoses, procedures, medications, and laboratory tests performed. Member hospitals represent 17 of the 20 major metropolitan areas across the United States, with only 1 children's hospital representing each city. On the basis of estimates from the National Association of Children's Hospitals and Related Institutions (Alexandria, VA), 70% of freestanding, pediatric, acute care hospitals in the United States submit data to the PHIS database.
Data quality and reliability are ensured through a joint effort between Child Health Corporation of America, a data manager (Thomson Healthcare, Evanston, IL), and participating hospitals. Data are deidentified at the time of submission and are subjected to a number of reliability and validity checks before being processed into data quality reports. Data are accepted into the database once classified errors occur in <2% of a hospital's quarterly data. Hospitals were excluded if they did not contribute laboratory data during the entire study period or if significant data problems were identified, leaving 22 hospitals in our study.
Children (<18 years of age) who were inpatients at PHIS hospitals between January 1, 2001, and December 31, 2006, were eligible for this study. Case status was determined by the presence of an International Classification of Diseases, Ninth Revision, Clinical Modification code for C difficile infection (code 008.45), a code that was described and validated previously.7 To improve the validity of our case definition, we included only cases with billing for a C difficile toxin assay and an initial dose of antimicrobial therapy directed against CDAD within the period of 1 day before to 2 days after the C difficile toxin assay. CDAD antimicrobial therapy was defined as metronidazole administered either orally or parenterally or vancomycin administered orally.
We examined all cases for the presence of a procedure code for colectomy. We excluded cases with a procedure code for colectomy if there was a concurrent diagnostic code for necrotizing enterocolitis of a newborn, because this condition is often treated with bowel resection.
Determination of Underlying Medical Conditions
To assess concurrent chronic conditions, we used a previously validated diagnostic classification system for pediatric, complex, chronic conditions that is based on International Classification of Diseases, Ninth Revision, Clinical Modification codes. This classification system divides International Classification of Diseases, Ninth Revision, codes into 9 categories of cancer and noncancerous conditions that require special health services, that is, neuromuscular, cardiovascular, respiratory, renal, gastrointestinal, hematologic or immunologic, metabolic, malignancy, and genetic or other congenital defects.8
Summary statistics were constructed by using frequencies and proportions for categorical data elements and means and medians for continuous variables, with SAS 9.1.3 (SAS Institute, Cary, NC). The Wilcoxon rank-sum test for trends and the Spearman rank correlation were performed with Stata 9 (Stata, College Station, TX).
In 2001–2006, we identified 4895 patients with CDAD (Fig 1), in 8276 276 total patient-days. Demographic characteristics are shown in Table 1. Boys accounted for 54% of the total number of cases, and 76% of patients were white. Four percent of the patients died during hospitalization. The median age of the children with CDAD was 4 years (interquartile range [IQR]: 1–11 years). There were 1267 patients (26%) ≤1 year of age and 249 (5%) 0 to 1 month of age.
From 2001 to 2006, the annual incidence density of CDAD increased from 2.6 to 4.0 cases per 1000 admissions (53%; P = .04), and the incidence rate increased from 4.4 to 6.5 cases per 10 000 patient-days (47%; P = .06) (Fig 2). When these trends were stratified according to center, 1 hospital had increases from 3.6 to 11.8 cases per 1000 admissions (227%) and from 7.1 to 23.1 cases per 10 000 patient-days (225%). When that center was excluded, the incidence densities of CDAD were 2.5 and 3.7 cases per 1000 admissions in 2001 and 2006, respectively (48% increase; P = .06). The incidence rates were 4.2 and 5.7 cases per 10 000 patient-days in 2001 and 2006, respectively (36% increase; P = .06). We segregated the 22 hospitals according to US Census regions to assess whether geography affected trends in CDAD incidence. No differences according to US Census region were observed.
We examined the possibility that increased testing explained the increased incidence of CDAD. Rates of testing for C difficile toxin did not increase from 2002 to 2006. Data on testing were not available for 2001.
When results were stratified according to age, the rates of CDAD among hospitalized children 1 to 5 years of age increased markedly (from 0.7 to 1.3 cases per 1000 admissions, 85%; P = .04; from 1.2 to 2.1 cases per 10 000 patient-days, 75%; P = .04). Similarly, the rates of CDAD among hospitalized children 5 to 17 years of age increased significantly (from 1.2 to 1.8 cases per 1000 admissions, 50%; P = .03; from 2.0 to 2.9 cases per 10 000 patient-days, 45%; P = .09). For children <1 year of age, we did not observe a significant difference in the incidence of CDAD over the study period (from 3.1 to 3.0 cases per 1000 admissions). In a secondary analysis, we excluded infants <1 year of age from the analysis, because it is unclear whether the presence of C difficile toxin represents true disease. The exclusion of infants did not significantly change our findings of an increased incidence of CDAD over time (P = .05).
In the cohort, 3293 children (67%) had ≥1 complex chronic condition; however, complex chronic conditions varied significantly across age groups (P < .0001) (Table 2). Cardiovascular conditions were most common in the youngest age group (0–1 month) and became less common with increasing age. Malignancy was the most common complex chronic condition in the oldest age group.
The majority of patients with CDAD (61%) were treated with single-therapy, orally administered metronidazole during their hospital stays. Oral metronidazole treatment increased significantly over time (P = .04), mirroring the overall increase in CDAD. Fewer than 5% of patients (3.5%) received orally administered vancomycin, and its use did not increase during the study period. All 172 patients who received only orally administered vancomycin were admitted at 5 (22.7%) of the 22 hospitals. Approximately 9% of patients received both vancomycin (administered orally) and metronidazole (administered orally and/or parentally), but not concurrently.
In our cohort, 61 patients underwent a colectomy during hospitalization. The median time for initiation of CDAD antimicrobial therapy was 3 days before colectomy (IQR: 14 days before to 4 days after colectomy). There was no significant increase in the incidence of colectomy for patients with CDAD during the study period. Children who underwent colectomy had a median age of 2.1 years (IQR: 0.2–11.9 years), and most were male (64%). The median total length of stay for patients who underwent colectomy was 34 days (IQR: 21–89 days). Overall, the all-cause mortality rate among patients with CDAD was 4% and did not increase over time.
We are the first to report the increasing nationwide burden of CDAD among children hospitalized at freestanding pediatric hospitals. We found an increasing number of cases of CDAD in hospitalized children over time, especially in the 1- to 5-year and 5- to 17-year age groups. We also found that 26% of the patients in our cohort were <1 year of age and 5% were newborns. In contrast to studies with adults, the colectomy and mortality rates for children with CDAD did not increase.9,10 The majority of CDAD cases were treated with single-therapy, orally administered metronidazole; in addition, the rate of single-therapy, oral metronidazole treatment increased significantly over time. Most cases of CDAD involved children with complex chronic conditions.
The incidence of CDAD among hospitalized children increased significantly, by 53%, from 2.6 to 4.0 cases per 1000 admissions. A number of studies reported increases in the incidence and severity of CDAD in various health care settings, but those studies were either performed at adult centers or used administrative data sources that included small proportions of pediatric patients.9–12 Previous studies that examined the rates of either nosocomial or community-acquired CDAD in children were generally single-center studies and reported inconsistent results, either that CDAD was an important increasing cause of diarrhea or that CDAD was not a significant cause of diarrhea in children.5,13,14 Our study was multicenter and included the largest number of pediatric CDAD cases reported.
There are many possible explanations for the increased number of CDAD cases. More individuals may be harboring C difficile, or there may be dissemination of a more-virulent epidemic strain of C difficile, such as North American pulsed-field gel electrophoresis type 1 (NAP1),15 which is considered a major factor for the recent increase in adults. The increased survival rates among chronically ill children might have increased the time at risk for acquiring nosocomial CDAD and the exposure to antimicrobial agents that promote the emergence of CDAD. In addition, coding for CDAD might have improved, or there might have been increased testing and treatment for CDAD during the study period.
C difficile has been seen traditionally as nonpathogenic in young infants and children.16,17 However, we found that 26% of CDAD cases occurred in children <1 year of age and 5% of all cases occurred in infants <1 month of age. Therefore, a significant proportion of children who received CDAD-directed antimicrobial therapy were at an age previously thought to be unaffected by C difficile toxin. One reason for our finding of CDAD in infants may be related to the emergence of the NAP1 strain; however, this remains to be proved. Severe CDAD caused by the NAP1 strain has been reported for persons previously thought to be at low risk, including children without risk factors for health care-associated CDAD.10,11 Another study reported that, of 22 NICU patients with CDAD, 20 had gastrointestinal pathologic conditions.13 Although our study could not examine the clinical course of each patient, our case definition included a treatment course of antimicrobial agents directed against CDAD, leading to 2 possible conclusions, that is, (1) CDAD affects young infants and newborns to a previously unrecognized degree or (2) a significant number of children are receiving CDAD-directed antimicrobial therapy inappropriately.
We did not find an increase in colectomy or mortality rates during the study period. Studies with adults have reported increased rates of both death and colectomy.10,11 CDAD caused by the NAP1 strain may be severe enough to necessitate therapy but not colectomy and may not cause death.
Orally administered metronidazole remained the most common antimicrobial agent used to treat CDAD in our cohort over the study period, and the rate of oral metronidazole treatment in cases of CDAD increased significantly over time. Some children (n = 172) in our cohort were treated only with orally administered vancomycin; this may represent a difference in practice at the 5 hospitals at which all of those patients were treated.
Another significant finding is that the majority of cases of CDAD (67%; P < .001) occurred in children with complex chronic conditions. The complex chronic conditions might have increased the risk for CDAD because of increased antibiotic exposure, longer hospitalizations, or more-frequent outpatient contact with health care settings. Also, the severity of underlying medical conditions may be a risk factor for CDAD in children, as has been reported for adults.18 The association noted in our study between CDAD and complex chronic conditions is generalizable to all freestanding children's hospitals because of the nature of the hospitals that participate in PHIS, that is, tertiary care pediatric centers to which children with complex chronic conditions are referred.
Use of the PHIS database offers the unique advantage of nation-level data on the incidence of CDAD in children. Administrative data sources such as PHIS are limited specifically with respect to the possibility of miscoded or inaccurate information. We minimized such potential biases by increasing the stringency of our case definition. Although a previous study with adult hospitals reported 78% sensitivity and 99.7% specificity of the International Classification of Diseases, Ninth Revision, Clinical Modification code for C difficile and laboratory-confirmed findings,7 we included the diagnostic code for C difficile, a charge for a C difficile toxin assay, and treatment for CDAD, to minimize potential miscoding and misclassification.
This is the first study to report the increase in CDAD among inpatients at children's hospitals. The rates of CDAD among hospitalized children increased significantly from 2001 to 2006, especially for children >1 year of age. Children with CDAD were more likely to have complex chronic conditions. The clinical and molecular epidemiological features of CDAD in children need to be further elucidated.
- Accepted March 13, 2008.
- Address correspondence to Jason Kim, MD, Department of Infectious Diseases, Children's Hospital of Philadelphia, 34th Street and Civic Center Boulevard, 9S28 South Tower, Philadelphia, PA 19104. E-mail:
The authors have indicated they have no financial relationships relevant to this article to disclose.
What's Known on This Subject
The incidence and severity of Clostridium difficile-associated disease in adults have increased in recent years, as reported in North America and Europe. It is unclear whether such trends are occurring in the pediatric population.
What This Study Adds
The incidence of Clostridium difficile-associated disease among hospitalized children increased from 2001 to 2006. This study is the largest, with data from 22 freestanding children's hospitals. The study also describes the epidemiological features of hospitalized children with C difficile-associated disease.
- ↵Kyne L, Hamel MB, Polavaram R, Kelly CP. Health care costs and mortality associated with nosocomial diarrhea due to Clostridium difficile. Clin Infect Dis.2002;34 (3):346– 353
- ↵Archibald LK, Banerjee SN, Jarvis WR. Secular trends in hospital-acquired Clostridium difficle disease in the United States, 1987–2001. J Infect Dis.2004;189 (9):1585– 1589
- ↵Klein EJ, Boster DR, Stapp JR, et al. Diarrhea etiology in a children's hospital emergency department: a prospective cohort study. Clin Infect Dis.2006;43 (7):807– 813
- ↵Feudtner C, Hays R, Haynes G, Geyer J, Neff J, Koepsell T. Deaths attributed to pediatric complex chronic conditions: national trends and implications for supportive care services. Pediatrics.2001;107 (6). Available at: www.pediatrics.org/cgi/content/full/107/6/e99
- ↵Pépin J, Valiquette L, Alary ME, et al. Clostridium difficile-associated diarrhea in a region of Quebec from 1991 to 2003: a changing pattern of disease severity. CMAJ.2004;171 (5):466– 472
- Copyright © 2008 by the American Academy of Pediatrics