Measles is an important cause of morbidity and mortality worldwide1 and outbreaks continue to occur in the United States, largely in unvaccinated patients.2 By the middle of 2019, measles cases topped over 835 in 23 states which represents the most cases seen in the United States in 25 years. We previously reported complications and resource utilization in 33 patients hospitalized for measles from 2011 to 2017 in a Midwestern children’s hospital.3 While that single-center study was informative for potential measles complications, we inferred that resource utilization might reflect local practice patterns and demographics. The objective of the current study was to describe demographics, complications and resource utilization for children hospitalized for measles among a national, convenience sample of US children’s hospitals.
MATERIALS AND METHODS
We conducted a retrospective, descriptive study of patients 0–18 years old admitted between January 1, 2004 and December 31, 2018 to a hospital contributing data to Pediatric Health Information Systems (PHIS) with an International Classification of Disease version 9 or 10 discharge diagnosis code for measles (055.X or B05.X). We identified complications based upon categorization of secondary discharge diagnosis codes into relevant clinical groupings. Complex chronic condition (CCC) flags were generated by PHIS using previously published coding categories including cardiovascular, congenital/genetic, gastrointestinal, hematologic/immunologic, malignant, metabolic, neurologic/neuromuscular, renal/urologic, respiratory, technology dependant and transplant-related conditions.4 We analyzed resource utilization excluding patients with CCCs,4 as it was not possible to distinguish measles utilization from other CCC indications. We used STATA version 14.0 (College Station, TX) for descriptive statistics. To avoid potential patient identification, outcomes aside from mortality with less than 5 patients were indicated as <5. This study was deemed exempt from review by the local Institutional Review Board.
Fifty-two hospitals reported on average 12 years of data to PHIS, and 34 hospitals reported at least 1 measles-related admission (median of 3 admissions per hospital, range 1–40). We identified 136 unique patients with 137 measles-related admissions. One patient had a measles-related readmission within 7 days of discharge. Median age was 23 months, 49% were male, 32% were white/Caucasian and 26% were black/African American, 57% had public insurance and 48% were from Midwestern hospitals. Twenty-three patients (17%) had CCCs, most commonly hematologic/immunologic conditions (7/23 or 30%), neurologic/neuromuscular (6/23, 26%) or technology dependency (6/23, 26%). Half of the patients identified with CCC (11/23) had more than 1 CCC. Two patients with CCC (2/23; 9%) died or were discharged to hospice during their measles-related admission.
Among patients without CCC, 53% were given at least 1 dose of antibiotics and 39% received Vitamin A. Dehydration (53%) and eye problems (25%) were the most common complications (Table 1). Among patients without CCC, median length of stay was 3 days (range 1–50 days) and median-adjusted, estimated hospital costs were $5896 (range $1089–$134,491).
Measles was a rare cause of hospitalization in 52 US children’s hospitals during the study period. Similar to previous reports, dehydration, pneumonia and tracheitis/croup were frequent complications. By contrast, keratitis/eye problems (25%) were higher than previously described.3
Two patients (9%) died or were discharged to hospice during this study period. A previous description of US measles cases during 1987–2000 showed a mortality rate of 0.3%.5 A separate estimate from the Centers for Disease Control and Prevention showed a mortality rate of 0.2% with most cases of death in children being caused by pneumonia.2 The higher mortality rate found in our study likely reflects the increased acuity in a hospitalized study population. Additionally, it may reflect higher risk of mortality and complications in patients with underlying CCCs, which were present in 17% of our study population. We previously reported a case series of 33 hospitalized patients in which no patients died. Notably no patients in that study had CCCs.3 Presence of CCCs may increase morbidity and rate of admission. One previous study of 248 children presenting to an Italian emergency department with measles found that those with CCC (13%) had higher odds ratio [7.95 (95% confidence interval 2.9–21.5)] of hospital admission than patients without CCCs.6
The majority (52%) of patients received at least 1 dose of antibiotics. We previously found in a single-center study that 73% (24/33) of patients received at least 1 dose of antibiotics.3 The difference in rate of antibiotic use may reflect differences in comorbid diagnoses [eg, otitis media rate 42% (14/33) vs. 27% in this study] as well as variation in antibiotic treatment practices across children’s hospitals. We were unable to identify the indication for antibiotic treatment because of the nature of the administrative database. A diagnosis of otitis media, pneumonia and sepsis were coded in 27%, 22% and 5% of patients, respectively, which would be potential antibiotic indications. In addition to treatment of identified bacterial coinfections, it is possible that some clinicians provided antibiotics empirically to prevent measles complications. A Cochrane review of antibiotics to prevent measles complications found a potential benefit to empiric treatment but did not provide specific recommendations for use because of the poor quality of existing studies.7
We found a low (39%) rate of vitamin A use across US children’s hospitals for measles similar to our previous single-center study with a rate of 21%.3 This may reflect differing recommendations for the use of vitamin A in measles,1,2,8 clinician confusion regarding recommendations for supplementation in developed versus developing countries, and a lack of comparative effectiveness studies in US patients. Future studies or expert panels should clarify the use of Vitamin A in US patients with measles. Additional questions or considerations include modification of the wording of Vitamin A recommendations to include specifications for different patient populations and regions, the role of Vitamin A in immune response to measles in different populations and clarifying if Vitamin A stores are depleted at the time of measles infection; thus, Vitamin A treatment is independent of underlying nutritional status.
Median-adjusted hospital cost per patient in our study was $5896, which is similar to the median cost of $5291 reported in our previous single-center study.3 This is higher than mean standardized cost per encounter estimates for common conditions such as asthma ($3799), dehydration ($3753) and infectious gastroenteritis ($4147) at US children’s hospitals.9 Furthermore, previous studies show costs to hospitals and health systems for managing cases of measles extend well beyond direct hospital costs3 reported in our study which did not include cost analysis for public health resource utilization in outbreak management and postexposure follow-up. One study estimated a public health cost of $20,000 per measles case.10
Study limitations are inherent to the retrospective nature and limitations to the PHIS database. PHIS does not capture immunization status, nutritional status, laboratory or imaging results and clinical reasoning for tests or treatment indications could not be determined. This database includes children’s hospitals and results may not be generalizable to community hospitals or other care settings.
Though it is an uncommon diagnosis at US children’s hospitals, patients hospitalized for measles have significant morbidity, mortality and inpatient resource utilization. Private and public health ventures should continue efforts to improve measles vaccination rates. Clinician education on recognition of measles and potential complications and treatment options is needed. Future studies are also needed to better understand the barriers and solutions to properly use Vitamin A and examine risks and benefits of Vitamin A and empiric antibiotic use in patients with measles in the United States.
3. Hester G, Nickel A, LeBlanc J, et al. Measles hospitalizations at a United States children’s hospital 2011–2017. Pediatr Infect Dis J. 2019;38:547–552.
4. Feudtner C, Feinstein JA, Zhong W, et al. Pediatric complex chronic conditions classification system version 2: updated for ICD-10 and complex medical technology dependence and transplantation. BMC Pediatr. 2014;14:199.
5. Orenstein WA, Perry RT, Halsey NA. The clinical significance of measles: a review. JID. 2004;189:S4–S16.
6. Ciofi Degli Atti M, Filia A, Bella A, et al. Measles cases in children requiring hospital access in an academic pediatric hospital in Italy, 2008-2013. Pediatr Infect Dis J. 2017;36:844–848.
7. Kabra SK, Lodha R. Antibiotics for preventing complications in children with measles. Cochrane Database Syst Rev. 2013:CD001477.
8. Huiming Y, Chaomin W, Meng M. Vitamin A
for treating measles in children. Cochrane Database Syst Rev. 2005:CD001479.
9. Keren R, Luan X, Localio R, et al; Pediatric Research in Inpatient Settings (PRIS) Network. Prioritization of comparative effectiveness research topics in hospital pediatrics. Arch Pediatr Adolesc Med. 2012;166:1155–1164.
10. Lo NC, Hotez PJ. Public health and economic consequences of vaccine hesitancy for measles in the United States. JAMA Pediatr. 2017;171:887–892.