ICD-9 coding unreliable source to determine primary RSV burden

VANCOUVER, Canada — ICD-9 codes often missed and did not identify respiratory syncytial virus disease in infants younger than 12 months in the ED during peak and shoulder seasons, according to data presented at the 2014 Pediatric Academic Societies Annual Meeting.

Doris Makari, MD, of MedImmune, and colleagues evaluated infants with lower respiratory infection or apnea who presented to 31 US EDs during two RSV seasons (2006 to 2008) to determine RSV prevalence assessed by real time-PCR compared with primary ICD-9 diagnosis. Peak RSV seasons from Jan. 15 to Feb. 15 and the seasons immediately before and after peak, or shoulder season, were from Sept. 1 to Oct. 31 and April 1 to May 31.

Doris Makari

Overall, discharge codes were available for 1,100 peak-season patients and 3,070 shoulder-season patients. Using real time-PCR (RT-PCR), 62% of peak-season patients and 20% of shoulder-season patients were positive for RSV. However, RSV-specific ICD-9 diagnosis codes were only assigned to 35% of peak-season patients and 25% of shoulder-season patients who tested positive for RSV using RT-PCR at discharge.

Acute bronchiolitis (50%) was the most common condition with higher RSV positivity rates among patients without an RSV-specific discharge diagnosis, followed by acute bronchiolitis due to other infectious organisms (33%); other dyspnea and respiratory abnormalities (30%); cough (24%); pneumonia (20%), wheezing (20%), acute upper respiratory illness of unspecified site (15%); and fever (12%).

“This study really shows that coding is not very good,” Marki told Infectious Diseases in Children. “It’s not a reliable source for databases when you’re really trying to capture the burden of disease for RSV. Physicians have to be very careful when just looking at ICD-9 codes because a lot of it will be nonspecific bronchiolitis when it’s a positive RSV; it may not have been coded correctly or they might not have done the test.” — by Amber Cox

For more information:

Makari D. Abstract 1522.333. Presented at: PAS 2014; May 2-6, 2014; Vancouver, Canada.

Disclosure: See the study for a full list of disclosures.