Fellow: Caleb Hancock, MD. University of Tennessee Health Science Center Memphis, Le Bonheur Children’s Hospital
Article: Desai S, Shah SS, Hall M, Richardson TE, Thomson JE. Imaging Strategies and Outcomes in Children Hospitalized with Cervical Lymphadenitis. J Hosp Med. 2020 April;15(4):197-203
Summary: As there are no national standards for imaging in cervical lymphadenitis, this study sought to characterize imaging practices in this area and evaluate for any associations between early imaging and outcomes. They performed a multicenter, cross-sectional study using data from the Pediatric Health Information Systems (PHIS) database including children ages 2 months to 18 years who were hospitalized between July 2013 and December 2017 with a diagnosis of cervical lymphadenitis. They looked at the proportion of patients that experienced any neck imaging, CT or ultrasound imaging, early imaging (defined as neck imaging conducted on day 0 of admission), and multiple imaging studies during a single hospitalization, and 10,014 patients were identified that met inclusion criteria. Early imaging was noted to be completed in 61% of children and 63.9% of these images were via ultrasound. Early imaging was associated with increased odds of receiving multiple imaging studies (adjusted odds ratio [aOR] : 3.0, 95% CI: 2.6-3.6), surgical drainage (aOR: 1.3, 95% CI: 1.1-1.4), 30-day readmission (aOR: 1.5, 95% CI: 1.2-1.9), and longer length of stay (adjusted rate ratio: 1.2, 95% CI 1.1-1.2). In a secondary analysis, when early imaging was redefined as imaging on day 0 or day 1, the odds of repeat imaging increased even more (aOR: 22.6, 95% CI: 15.8-32.4)
Key Strengths: By using the PHIS database, the authors were able to create a well-powered study with a large patient population representative of 44 distinct institutions. Verification of a structured algorithm to identify patients with a positive predictive value of 95.1% when ICD-10 codes were considered helped to ensure that patients were classified and captured appropriately for this study. It was also important that they excluded children from hospitals with fewer than 50 cases of cervical lymphadenitis as they could have skewed measures of hospital-level variation.
Limitations: Use of the PHIS database can also be a limitation as only administrative and billing data is available, so information such as exam findings or radiographic interpretations were not available to further characterize the lymphadenopathy. The database limited the ability to define early imaging (day 0) to only those occurring before midnight on the day of presentation rather than using a full 24-hour period. Additionally, they could not identify imaging that occurred prior to presentation to one of the hospitals represented in PHIS (such as at an outside emergency department, outpatient clinic, or urgent care), so some patients with early imaging could have been missed.
Main Takeaway: Children who received early imaging experienced more resource utilization and more interventions than those that did not.
Impact on Practice: This study provides evidence for overuse of imaging in cervical lymphadenitis. Care of these patients would likely benefit from standardization of care to help guide decision-making. Additionally, considering the natural history of cervical lymphadenitis, expectant management with antibiotics and follow up may be beneficial before considering radiological imaging leading to surgical drainage when necessary.
Article: Desai S, Shah SS, Hall M, Richardson TE, Thomson JE. Imaging Strategies and Outcomes in Children Hospitalized with Cervical Lymphadenitis. J Hosp Med. 2020 April;15(4):197-203
Summary: As there are no national standards for imaging in cervical lymphadenitis, this study sought to characterize imaging practices in this area and evaluate for any associations between early imaging and outcomes. They performed a multicenter, cross-sectional study using data from the Pediatric Health Information Systems (PHIS) database including children ages 2 months to 18 years who were hospitalized between July 2013 and December 2017 with a diagnosis of cervical lymphadenitis. They looked at the proportion of patients that experienced any neck imaging, CT or ultrasound imaging, early imaging (defined as neck imaging conducted on day 0 of admission), and multiple imaging studies during a single hospitalization, and 10,014 patients were identified that met inclusion criteria. Early imaging was noted to be completed in 61% of children and 63.9% of these images were via ultrasound. Early imaging was associated with increased odds of receiving multiple imaging studies (adjusted odds ratio [aOR] : 3.0, 95% CI: 2.6-3.6), surgical drainage (aOR: 1.3, 95% CI: 1.1-1.4), 30-day readmission (aOR: 1.5, 95% CI: 1.2-1.9), and longer length of stay (adjusted rate ratio: 1.2, 95% CI 1.1-1.2). In a secondary analysis, when early imaging was redefined as imaging on day 0 or day 1, the odds of repeat imaging increased even more (aOR: 22.6, 95% CI: 15.8-32.4)
Key Strengths: By using the PHIS database, the authors were able to create a well-powered study with a large patient population representative of 44 distinct institutions. Verification of a structured algorithm to identify patients with a positive predictive value of 95.1% when ICD-10 codes were considered helped to ensure that patients were classified and captured appropriately for this study. It was also important that they excluded children from hospitals with fewer than 50 cases of cervical lymphadenitis as they could have skewed measures of hospital-level variation.
Limitations: Use of the PHIS database can also be a limitation as only administrative and billing data is available, so information such as exam findings or radiographic interpretations were not available to further characterize the lymphadenopathy. The database limited the ability to define early imaging (day 0) to only those occurring before midnight on the day of presentation rather than using a full 24-hour period. Additionally, they could not identify imaging that occurred prior to presentation to one of the hospitals represented in PHIS (such as at an outside emergency department, outpatient clinic, or urgent care), so some patients with early imaging could have been missed.
Main Takeaway: Children who received early imaging experienced more resource utilization and more interventions than those that did not.
Impact on Practice: This study provides evidence for overuse of imaging in cervical lymphadenitis. Care of these patients would likely benefit from standardization of care to help guide decision-making. Additionally, considering the natural history of cervical lymphadenitis, expectant management with antibiotics and follow up may be beneficial before considering radiological imaging leading to surgical drainage when necessary.