key: cord-1027040-kmcgsmgt
authors: Bolge, Susan C.; Gutierrez, Cynthia; Kariburyo, Furaha; He, Ding
title: Burden of Pneumonia Among Hospitalized Patients with Influenza: Real-World Evidence from a US Managed Care Population
date: 2021-08-16
journal: Pulm Ther
DOI: 10.1007/s41030-021-00169-2
sha: 18884b4bcdbecb6ada4e7737b04ae2d0620ad978
doc_id: 1027040
cord_uid: kmcgsmgt

INTRODUCTION: Pneumonia is among the most prevalent complications of influenza. The purpose of this study is to quantify the burden of pneumonia among hospitalized patients with influenza. METHODS: Real-world retrospective data from 01JAN2014-30JUN2019 (study period) were obtained from Optum’s de-identified Clinformatics® Data Mart Database (2007–2020) for patients who had ≥ 1 diagnosis for influenza during the identification period and ≥ 1 all-cause inpatient visit within 1 day of diagnosis. Cases had ≥ 1 diagnosis claim for an influenza-related pneumonia within the 30 days after the initial influenza diagnosis date. Controls had no evidence of influenza-related pneumonia in the 30 days following the initial influenza diagnosis. Final 1:1 matching was determined using propensity score matching (PSM). Statistical significance between the cohorts was tested. RESULTS: After PSM, there were 4878 hospitalized patients with influenza in each of the case and control groups. During the index hospitalization, cases vs. controls had longer length of stay [Mean (standard deviation): 6.5 (8.3) vs. 1.9 (3.7)], greater intensive care unit (ICU) use (38.4 vs. 16.8%), and greater mechanical ventilation use (invasive: 11.4 vs. 2.3%; non-invasive: 6.8 vs. 2.6%) (all p < 0.001). Cases also had higher readmission rates than controls (12.3 vs. 3.5% within 30 days; 20.0 vs. 6.1% within 90 days; p < 0.001 for both). Post-index date direct all-cause healthcare costs were higher for cases than for controls (median total cost: $18,428 vs. $621 for 30 days; $21,774 vs. $3312 for 90 days; $25,960 vs. $8699 for 6 months; $35,875 vs. $21,619 for 1 year; all p < 0.001). CONCLUSIONS: Pneumonia as a complication of influenza increases risk of mortality and leads to greater healthcare resource use and direct medical costs among patients hospitalized with influenza. These effects are seen early during the index hospitalization and within the first 30 days after diagnosis, but their impact continues throughout a year of follow-up. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1007/s41030-021-00169-2.

Influenza is an acute respiratory infection, having affected 9.2 million to 35.6 million people in the United States from the 2010-2011 to the 2015-2016 influenza season, resulting in increased morbidity and mortality [1] . Most of the resulting morbidity and mortality in influenza patients is due to complications disproportionately affecting vulnerable groups such as individuals with chronic disease (including chronic pulmonary disease), individuals aged C 50 years, pregnant women, children aged 6-59 months, individuals receiving aspirin-or salicylate-containing medications, extremely obese individuals, and immunecompromised individuals [2, 3] . Pneumonia, either bacterial or viral, is the most frequent serious complication and usually occurs towards the resolution of influenza symptoms 4-14 days after diagnosis [4, 5] . Together, influenza and pneumonia are the eighth leading cause of death in the United States [6] . Respiratory-related deaths from the 2010-2011 to the 2013-2014 influenza season ranged from 12,000 to 56,000. Pneumonia-associated and influenza deaths ranged from 4000-12,000 between the 2010-2011 and the 2015-2016 influenza season [1] .

Complications from influenza resulting in hospitalization remain a significant health concern. During the 2017-2018 influenza season, the US Centers for Disease Control and Prevention (CDC) estimates there were [ 45 million cases of influenza resulting in [ 21 million influenza-associated medical visits, [ 810,000 influenza-related hospitalizations, and [ 61,000 deaths [7] . Along with the physical burden of influenza, there is also a significant economic burden. Previously, the projected annual cost of influenza-related healthcare has been estimated to range from $2.0-$5.8 billion (for two of the most common sub-types) to $10.4 billion [8, 9] .

Previous studies either examined influenzarelated or pneumonia-related healthcare resource utilization (HCRU). Many of the studies attempting to quantify the burden of influenza do not calculate the burden of influenzaassociated pneumonia [10] [11] [12] [13] [14] [15] [16] . At the same time, many pneumonia studies evaluate the burden of pneumonia (not including those evaluating viral or bacterial pneumonia) regardless of the etiology [17, 18] . More specifically, there is a lack of real-world studies focusing on the burden of pneumonia as a complication of influenza-the most frequently reported complication of influenza. The purpose of this study was to quantify the burden of pneumonia among hospitalized patients with influenza.

This real-world observational study was conducted using the Optum de-identified Clinfor-maticsÒ Data Mart Database (2007-2020) (''Optum'') from January 2014 through June 2019. The Optum database captures administrative claims for [ 33 million privately insured members with both medical and pharmacy benefits.

Optum claims data files that were used for this study included medical claims, pharmacy claims, and laboratory results. All patient identifiers in the database have been fully encrypted, and the database is fully compliant with the Health Insurance Portability and Accountability Act (HIPAA) of 1996. As such, Institutional Review Board (IRB) approval to conduct this study was not required and considered exempt according to 45CFR46.101(b)(4): Existing Data & Specimens-No Identifiers.

Patients with a medical claim for influenza diagnosis (International Classification of Disease, 9th Revision, Clinical Modification [ICD-9-CM] codes 487, 488; ICD-10-CM codes J09, J10, J11) were identified between January 1, 2015 and June 30, 2018. The date of the first observed influenza diagnosis during this period was designated as the initial influenza diagnosis date. Patients were required to have C 1 allcause inpatient visit within 1 day of the initial influenza diagnosis date during the identification period. The first inpatient visit during this period was designated as the initial influenza hospitalization.

Patients were included in the case cohort if they had C 1 diagnosis for an influenza-related pneumonia (ICD-9-CM: 487.0, 488.01, 488.11, 488.81; ICD-10-CM: J09.X1, J10.0x, J11.0x) within the 30 days after the initial influenza diagnosis date during the identification period (01JAN2015-30JUN2019). The first influenzarelated pneumonia diagnosis date was designated as the index date. Control patients were identified as patients with no evidence of influenza-related pneumonia in the 30 days following the initial influenza diagnosis. For each case patient, a control patient of identical age, sex, and region was selected during the same index study year. To reduce bias, distribution of time between initial influenza diagnoses and influenza-related pneumonia diagnoses for cases was calculated. The index date for controls was then randomly assigned based on the distribution of influenza-related pneumonia diagnoses to the index date for case patients.

Patients in both case and control cohorts were required to be C 13 years at the time of index, with continuous medical and pharmacy benefits for C 12 months prior to the index date (baseline period) and C 12 months after the index date (follow-up period). Patients who died during the follow-up period were also included in the study. To exclude patients with recurrent pneumonia leading up to the initial influenza diagnosis, patients were excluded if they had a diagnosis of pneumonia in the 30 days before the initial influenza diagnosis. Patients with recurrent pneumonia during the full baseline period were not excluded. (Fig. 1) 

Baseline patient demographics including age, sex, US geographic region, health insurance plan type, and Charlson Comorbidity Index (CCI) score were evaluated [19, 20] . Baseline comorbidities were assessed during the 12-month baseline period for all patients. Comorbidities of interest included ankylosing spondylitis, anxiety, asthma, cancer, chronic kidney disease, chronic liver disease, chronic obstructive pulmonary disease, congenital heart disease, congestive heart failure, coronary artery disease, Crohn's disease, cystic fibrosis, depression, diabetes, leukemia, high cholesterol, HIV, hypertension, lymphoma, multiple sclerosis, osteoarthritis, osteoporosis, plaque psoriasis, plasma cell neoplasms psoriatic arthritis, rheumatoid arthritis, solid organ transplant, stem cell transplant, stroke, systemic lupus erythematosus, and ulcerative colitis. All comorbid conditions were identified using ICD-9-CM or ICD-10-CM codes.

Characteristics of the initial influenza hospitalization were examined including the length of stay (LOS), intensive care unit (ICU) use, mechanical ventilation (MV) use, and initial influenza hospitalization healthcare costs. In addition, cumulative 30-and 90-day readmission as well as all-cause mortality during 30 days, 90 days, 6 months, and 12 months in the follow-up period were examined. All-cause HCRU and costs during the pre-and post-index period were assessed, including hospital LOS, office visits, pharmacy use, emergency department/room (ER), inpatient (across all hospitalization), and cumulative outpatient visits during 30 days, 90 days, 6 months, and 12 months in the follow-up. Costs were adjusted to 2019 US dollars using the medical care component of the Consumer Price Index.

All baseline and outcome variables were analyzed descriptively and compared between cases and controls. Propensity score matching (PSM) was performed to adjust for observed differences between the case and control cohorts. An unconditional logistic regression was fitted to determine the baseline characteristics associated with being in the case or control cohort. Using this model, a propensity score was developed for each patient which characterized the probability of being a member of the case cohort. Covariates in the propensity model included baseline demographic and clinical characteristics such as age, sex, US geographic region, health plan, payer type, pre-index CCI score, baseline all-cause HCRU, and individual comorbidities. Each patient in the case cohort was matched to a patient in the control cohort with the closest propensity score. Percentages and counts were provided for categorical variables. Means and standard deviations (SDs) were 

There was a total of 106,914 patients aged C 13 years old with a diagnosis of influenza during the identification period (01JAN2015-30JUN2018). A total of 51,638 (48%) patients were hospitalized within 0--1 days of being diagnosed with influenza in the Optum database. After applying the inclusion and exclusion criteria, the final analytical sample was composed of a total of 25,159 patients including 12,350 influenza-related pneumonia cases (49.1%) and 12,809 potential controls (50.9%) without an influenza-related pneumonia diagnosis in the 30 days after the initial influenza diagnosis. After performing a 1:1 hard matching ratio based on age, gender, and region, there was a total of 11,169 patients with influenza-related pneumonia cases and 11,169 controls.

Pre-PSM patient characteristics and baseline HCRU and costs are summarized in Table 1 . Prior to PSM, the average age among influenza- (Table 1) .

The characteristics of patients' initial influenza hospitalization are presented in Table 2 . Compared to controls, influenza-related pneumonia cases had longer initial influenza hospitalization (6.5 days vs. 

Mortality and hospital readmission trends are presented in Table 3 . Compared with controls, influenza-related pneumonia case patients also had higher mortality rates at 30 days ( 

This real-world retrospective data study quantified the burden of pneumonia among hospitalized patients with influenza using a large US managed care population. Overall, 48% of influenza patients were initially hospitalized within 0-1 day of diagnosis. After controlling for baseline differences using PSM, the baseline characteristics were well balanced between hospitalized influenza patients with influenzarelated pneumonia (cases) versus influenza patients without (controls).

The findings of this study suggest that pneumonia as a complication of influenza increases the risk of mortality and hospital readmission and also leads to greater HCRU and direct medical costs among patients hospitalized with influenza. The impact of pneumonia on these outcomes was evident particularly during the index hospitalization and within the first 30 days after diagnosis; however, their effects continued during the course of the first year of follow-up.

In most influenza cases, the infection is a self-limiting disease that resolves on its own without the risk of developing serious complications [21] [22] [23] . However, most of the resulting morbidity and mortality in influenza patients is due to complications, with pneumonia being the most prevalent complication of influenza [5, 24] . The average age of this study population was 76 years, which is consistent with previous findings that adults aged C 65 years are at high risk of developing pneumonia and being hospitalized [5, 10, 25] . This risk, as well as the risk for subsequent reinfection and slow recovery, may remain elevated for several months. The current guidelines recommend both influenza and pneumococcal vaccines in patients aged [ 65 years [26] . However, the uptake for both vaccinations is low in this high-risk population [26] . The results of this study underscore the need to prevent and reduce the risk of pneumonia in the elderly population alongside the associated burden in mortality and morbidity.

During the index hospitalization, case patients with influenza-related pneumonia remained in the hospital 5 days longer than controls; they were also more likely to be in the ICU and require MV and incurred more than $13,743 compared to controls. These findings are consistent with studies that have demonstrated that the duration of influenza associated hospitalization averages 4-8 days: this is dependent of several factors including patient's age, the nature of complications and comorbidities, and whether patients require the use of MV, which may impact hospital LOS [27] [28] [29] [30] . These factors potentially lead to increased economic burden.

In our study, we found significant 30-day readmission percentage differences between our cases and controls. Similar work calculating 30-day readmission among Canadian patients hospitalized with acute respiratory distress with laboratory-confirmed influenza calculated their 30-day readmission to be 5.7% [31] . While our results seem different, the 30-day readmission of our pooled population is 7.9%. The difference we are observing in our cohorts may be representative of the spectrum of influenza patients hospitalized and then readmitted.

It is known that influenza-related pneumonia has long been associated with increased morbidity and mortality [25, 29, 32] . In addition, deaths caused by influenza are usually due to complications including secondary pneumonia or worsening of an underlying condition. Previous studies have reported varying mortality rates from 11.78 per 100,000 personyears in the general population to 100 per 100,000 person-years in the Veterans Affairs population [15, 25] . Mortality rates as high as 141.15 per 100,000 person-years have been noted for older adults aged C 75 compared to adults aged 65-74 [10, 25] . Between October 2017 and January 2018, the weekly percentage of deaths attributed to pneumonia and influenza has ranged from 5.8 to 10.1% [33] . Both cases and controls cohorts in this study were elderly hospitalized influenza patients with a Previous studies have also focused exclusively on patients with influenza-related complications [4, 9, 34] . Our results were consistent with those of a previous study in which administrative claims were also used to analyze healthcare costs and HCRU between influenza patients with and without influenza-related complications [4] . Both healthcare costs and HCRU were found to be higher among influenza patients with complications and critical illnesses necessitating hospitalization [4] . Among influenza patients in large commercial claims database, Karve et al. estimated mean predicted total healthcare costs to be twice as high for influenza patients with influenza-related complications versus those without ($3422 vs. $1510) at the end of the 1-year follow-up period [4] . Our study found differences in costs between cases versus controls; however, our estimated costs were significantly higher for both cohorts ($64,052 vs. $43,976). It should be noted that Karve et al. did not require influenza patients to have an inpatient visit and did not restrict the timeframe for the outcome of the index influenza-related complication [4] . Additionally, their list of influenza-related complications was much more inclusive than ours and included various other conditions including acute otitis media, cardiovascular disease, and respiratory disorders [4] .

Putri et al. focused on estimating the economic burden of influenza regardless of the incidence of complications [34] . The authors used previously published health outcome rates and age-specific symptomatic attack rates for seasonal influenza to estimate the direct and indirect costs of influenza, including the costs of hospitalizations among patients with specified influenza ICD codes [34] . Their mean estimated costs for hospitalized patients aged 65-84 (mean: $8330) falls in between the mean costs of both cohorts' index hospitalization cost (mean: $18,221 vs. $4478) [34] . Subsequent costs for our case patients are higher, as these are the patients that developed pneumoniarequiring further HCRU.

To our knowledge, this is the largest retrospective observational study to compare the risk of ICU admission, mechanical ventilation use, hospital re-admission, mortality and economic burden in patients with a pneumonia diagnosis within 30 days of a hospitalization for influenza compared to patients who did not have a diagnosis of pneumonia in the 30 days after influenza hospitalization. There is a need for early and appropriate management of hospitalized influenza patients with influenza-related pneumonia. The findings from this study have demonstrated that complications resulting from influenza infection-particularly pneumoniacan lead to increased morbidity including longer hospital stay, ICU and MV utilization, risk of mortality, and increased HCRU and costs, especially in the elderly population. The need for early and appropriate management to avoid influenza-related complications has become even more emergent during the COVID-19 pandemic, which may place influenza patients in competition with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) patients for already scarce hospital resources.

Our study had multiple limitations that need to be taken into consideration. While claims data are extremely valuable for the efficient and effective examination of healthcare outcomes, resource use, and costs, claims data are collected for the purpose of payment and not research. Therefore, certain limitations are associated with claims data use. Given the observational nature of the study, only associations without causal linkage can be inferred. Pneumonia diagnoses were captured based on ICD-9/ICD-10 codes and not based on clinical parameters. The presence of an ICD-9/ICD-10 diagnosis code for pneumonia on a medical claim does not necessarily indicate the presence of the condition which may have resulted in patient misclassification. Influenza diagnosis was also identified using ICD-9/ICD-10 codes as opposed to laboratory data which could have led to the inclusion of influenza-like illness. Retrospective claims analyses are subject to coding errors or incorrectly entered diagnoses that were primarily coded for reimbursement purposes rather than clinical accuracy, therefore, some medical information may be unavailable or inaccurate.

Influenza-related pneumonia among those hospitalized with influenza led to an increased risk of mortality and greater HCRU and direct medical costs. These effects were seen early during the index hospitalization in the form of longer LOS and greater MV and ICU use and continued within the first 30 days after influenza diagnosis, but their impact remained throughout a year of follow-up. These findings demonstrate the need for treatments that reduce influenza complications-especially during the index hospitalization and in the 30 days after an influenza diagnosis.

Authorship. All authors had full access to all of the data in this study and take complete responsibility for the integrity of the data and accuracy of the data analysis. All named authors meet the International Committee of Medical Journal Editors (ICMJE) criteria for authorship for this article, take responsibility for the integrity of the work as a whole, and have given their approval for this version to be published.

Author Contributions. All authors contributed to the concept and design, statistical analysis, and drafting the manuscript.

Medical Writing, Editorial, and Other Assistance. Editorial assistance was provided by Chris Haddlesey and Michael Moriarty who are paid employees of SIMR, LLC, which is a paid consultant to Janssen Global Services, LLC.

Disclosures. Susan C. Bolge is an employee of Janssen Global Services, LLC, the study sponsor. Cynthia Gutierrez is an employee of SIMR, LLC, a paid consultant to the study sponsor. Furaha Kariburyo and Ding He were paid employees of SIMR, LLC at the time of study.

Compliance with Ethics Guidelines. Optum claims data files that were used for this study included medical claims, pharmacy claims, and laboratory results. All patient identifiers in the database have been fully encrypted, and the database is fully compliant with the Health Insurance Portability and Accountability Act (HIPAA) of 1996. As such, IRB approval to conduct this study was not required and considered exempt according to 45CFR46.101(b)(4): Existing Data & Specimens-No Identifiers.

Data Availability. The raw data on which this analysis was based are not publicly available due to a data licensing agreement.

Open Access. This article is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License, which permits any non-commercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/bync/4.0/.

Annual estimates of the burden of seasonal influenza in the United States: a tool for strengthening influenza surveillance and preparedness. Influenza Other Respir Viruses

The burden of influenza: a complex problem

Prevention and control of seasonal influenza with vaccines: recommendations of the advisory committee on immunization practices-United States, 2019-20 influenza season

Annual all-cause healthcare costs among influenza patients with and without influenza-related complications: analysis of a United States managed care database

Complications of seasonal and pandemic influenza

Deaths: leading causes for 2017

Update: influenza activity in the United States during the 2017-18 season and composition of the 2018-19 influenza vaccine

The annual impact of seasonal influenza in the US: measuring disease burden and costs. Vaccine

US healthcare costs attributable to type A and type B influenza

Age-related differences in hospitalization rates, clinical presentation, and outcomes among older adults hospitalized with influenza-U.S. Influenza Hospitalization Surveillance Network (FluSurv-NET)

Health care resource utilization and costs for influenza-like illness among midwestern health plan members

Epidemiology and burden of influenza in the U.S. Department of Veterans Affairs. Influenza Other Respir Viruses

Influenza-associated hospitalizations by industry, 2009-10 influenza season. United States Emerg Infect Dis

Estimating influenza disease burden from population-based surveillance data in the United States

The annual burden of seasonal influenza in the US Veterans Affairs population

Estimated rates of influenza-associated outpatient visits during 2001-2010 in 6 US integrated healthcare delivery organizations. Influenza Other Respir Viruses

Pneumonia-associated hospitalizations

Burden of pneumonia-associated hospitalizations: United States

Coding algorithms for defining comorbidities in ICD-9-CM and ICD-10 administrative data

The Charlson comorbidity index is adapted to predict costs of chronic disease in primary care patients

The burden of influenza complications in different high-risk groups: a targeted literature review

People at high risk for flu complications. Centers for Disease Control and Prevention website

Complications of viral influenza

Age-and sex-related risk factors for influenza-associated mortality in the United States between 1997-2007

Management of community-acquired pneumonia in older adults

Influenza-associated hospitalizations in the United States

Factors associated with early hospital discharge of adult influenza patients

Outcomes of adults hospitalised with severe influenza

Diagnosis, management and outcomes of adults hospitalized with influenza

Resource utilization and cost of influenza requiring hospitalization in Canadian adults: a study from the serious outcomes surveillance network of the Canadian immunization research network. Influenza Other Respir Viruses

Estimating the burden of seasonal influenza in Spain from surveillance of mild and severe influenza disease

Update: influenza activity-United States

Economic burden of seasonal influenza in the United States

Funding. This study was funded by Janssen Global Services, LLC. Rapid Service Fees were funded by Janssen Global Services, LLC.