2 tables omitted
The first cases of 2009 pandemic influenza A (H1N1) in New York City occurred in April 2009, raising many questions about how best to contain the epidemic. To rapidly assess the severity of influenza illness and identify persons at highest risk for severe infection, the New York City (NYC) Department of Health and Mental Hygiene (DOHMH) reviewed the medical charts of the first 99 patients with laboratory confirmed H1N1 admitted to any NYC hospital. The purpose of the review was to characterize the demographics of the first hospitalized patients, identify associated underlying medical conditions, describe the course and severity of disease, and examine the use of antiviral medications. This report summarizes the findings of this analysis. Approximately 60% of admitted patients were aged <18 years. The most commonly documented underlying condition was asthma, observed among 50% of patients aged <18 years and 46% of adult patients. Multiple underlying conditions were observed in 17% of patients (12% of children, 24% of adults). Patients treated with oseltamivir within 2 days of symptom onset had shorter median hospitalizations than those who did not (2 days versus 3 days [p = 0.03]). The findings of this assessment were used to inform immediate outbreak response measures in New York City. During such outbreaks, public education campaigns should encourage patients at high risk of severe illness to seek treatment promptly after symptom onset and should emphasize the importance of early antiviral therapy for patients with underlying risk conditions.1,2
The subjects of the assessment were the first 99 patients with polymerase chain reaction-confirmed H1N1 influenza admitted to any NYC hospital during April 25–May 24, 2009. To conduct the assessment, DOHMH physicians used a modified abstraction form based on one developed by CDC to collect clinical and laboratory data from paper and electronic medical charts of the hospitalized patients. Reviewing physicians identified underlying conditions known to increase risk for severe influenza.1,2 Body mass index (BMI) was calculated using height and weight recorded in the chart; BMI percentile-for-age for patients aged 2-17 years was determined by using CDC growth charts3 and the standard formula4 was used for nonpregnant adults aged ≥18 years. Patients with BMI ≥30 were categorized as obese.4 Wilcoxon ranked-sum tests were conducted to compare median lengths of hospitalization among surviving patients (statistically significant results defined as p< 0.05).
Among the 99 hospitalized patients, 19 (19%) were aged <5 years, 39 (39%) were aged 5-17 years, and nine (9%) were aged ≥50 years. These proportions differed from the proportions for the same age groups in the general population (2007 census projections for New York City), which were 7%, 16%, and 29%, respectively, indicating that hospitalized patients were generally younger than the general population. Of the hospitalized patients, non-Hispanic Asians and whites were underrepresented compared with the NYC general population, and Hispanics were overrepresented.
The most common presenting symptoms were fever and cough. A total of 95 patients (96%) had measured or subjective fever on admission; 56 (57%) had measured fever of >100.4°F (median maximum temperature: 102.2°F [39.0°C]; range: 97.0°F-105.9°F [36.1°C-41.1°C]) and 39 (39%) had subjective fever. A total of 89 (90%) reported cough. Additional presenting symptoms reported included runny nose (42%), shortness of breath (34%), headache (33%), vomiting (32%), and myalgias (31%). Elevated heart and respiratory rates for age were observed in 63 (64%) and 48 (48%) of patients, respectively. Abnormally high (20 patients) and low (seven patients) white blood cell counts were observed in 27 patients (27%). A total of 87 patients (88%) received at least one chest radiograph, of which 38 (44%) were read as abnormal. The most common abnormalities were single lower lobe infiltrates (24%), interstitial infiltrates (18%), and multilobar infiltrate (8%). Complications observed during hospitalizations included acute respiratory distress syndrome (ARDS) in three patients (3%), shock in three (3%), sepsis in five (5%), liver impairment in five (5%), and renal failure in five (5%).
Underlying medical conditions known to increase the risk of severe influenza or influenza complications1,2 were observed in 73 patients (74%), including 37 children (64%) and 36 adults (88%), and 17 patients (17%), including seven children (12%) and 10 adults (24%), had more than one underlying condition. The most commonly documented underlying condition was history of asthma, recorded for 29 patients aged <18 years (50%) and 19 adults (46%). Also recorded were chronic metabolic disorders including diabetes (11 patients [11%]), neurological disorders including neuromuscular disorders, seizure disorders, or cognitive dysfunction (10 patients [10%]), and immunosuppressive conditions, including HIV or medication-related conditions (five patients [5%]). Among the 24 female patients aged 15-49 years at the time of hospital admission, seven (29%) were pregnant or within 10 days after delivery, of whom four had additional underlying conditions. Among the 20 adults and 28 patients aged <18 years for whom information was available, 12 adults (60%) and five patients aged <18 years (18%) were obese. Underlying conditions1,2 were observed in 11 of the obese adults and four of the obese patients aged <18 years.
Among 24 patients (24%) admitted to the intensive care unit (ICU), seven (29%) required mechanical ventilation. Median age of ICU patients was 19 years (range: 0-55 years). Patients admitted to the ICU had longer median lengths of stay (4 days, range: 1-29 days) compared with other hospitalized patients.
Four patients (4%) died. Three of those patients were obese. Underlying conditions among the four included asthma (two) and Down syndrome (one). One patient died on the day of admission, two other patients died within 4 days of admission, and the fourth patient died 41 days after admission.
Median length of time from symptom onset to admission was 2 days (range: 0-14). Among the 95 patients who survived their hospitalization, a difference of 1 day for median length of hospitalization was observed for children compared with adults (2 days [range: 0-20] vs. 3 days [range: 1-29]; p = 0.01).
Antiviral treatment with oseltamivir was received by 76 patients (77%); three (4%) initiated treatment before hospitalization. Of the 76 patients who received antivirals, 36 (47%) began treatment within 2 days of symptom onset. Median time from onset of illness to treatment was 3 days (range: <24 hrs to 15 days). Patients who initiated antiviral treatment within the 2 days recommended by CDC1 had shorter lengths of stay than those who initiated treatment later (median: 2 days versus 3 days; p = 0.03).
ME Slopen, MSW, MC Mosquera, MD, S Balter, MD, BD Kerker, PhD, MA Marx, PhD, MR Pfeiffer, MPH, A Fine, MD, J Eavey, MSPH, TG Harris, PhD, EC Olson, MPH, C Stayton, PhD, C Wu, MPH, EH Lee, MD, New York City Dept of Health and Mental Hygiene.
This review was conducted to assess rapidly, in the first days of the NYC H1N1 outbreak, the characteristics and severity of illness in hospitalized patients in New York City. All data were collected within 2 weeks and results were available quickly to DOHMH to inform outbreak response measures. A key finding was that the first hospitalized patients in New York City were in younger age groups (91% of these patients were aged <50 years, and 59% were aged <18 years). Only one patient was aged ≥65 years, the most commonly hospitalized group for seasonal influenza.2 These findings were consistent with other descriptions of hospitalized persons with H1N1,5 and contributed to a growing body of national epidemiologic data that later informed ACIP recommendations6 regarding target groups for the forthcoming monovalent H1N1 vaccine.
The finding that prompt treatment with antiviral medications was associated with a shorter length of hospitalization did not definitively establish that more rapid treatment resulted in shorter hospitalization, in part because of a small sample size and possible confounding with other risk factors. However, the finding, along with CDC guidance concerning treatment of H1N1 influenza, generally supported DOHMH's public health messages that persons with underlying conditions should seek care as early as possible. This message was disseminated via press releases to the public, information published on the DOHMH website, and through electronic health alerts sent to NYC health-care providers.
Asthma was the most commonly noted underlying condition among H1N1 patients, observed for 50% of patients aged <18 years and 49% of adult patients (age adjusted). These proportions are higher than rates of asthma among NYC residents, as reported in the 2003 and 2007 NYC Community Health Survey (CHS),7 in which 17% of children aged <18 years and 13% of adults were reported to have a history of asthma. Although the ascertainment methods for asthma history were different for the hospital assessment and the CHS, the finding suggested that asthma might be playing a role in the earliest hospitalizations for H1N1 influenza. Similarly, obesity was more common among H1N1 patients (56% of adults [age adjusted] where information was available) than NYC residents (22% of adult NYC residents on the 2007 CHS).
Although BMI was available for only 59% of patients aged >2 years, 92% of obese adults and 80% of obese patients aged <18 years had an underlying condition, potentially increasing the risk for severe influenza or complications. In addition, three of the four fatalities occurred in obese individuals. Whether obesity itself contributes to the risk of acquiring H1N1 influenza or to the risk of severe disease or death remains unclear but has been a focus of investigation during the H1N1 pandemic.8,9 To aid in future studies, all patients hospitalized with H1N1 influenza should have an objective measure of height and weight documented in their medical record.
The findings in the report are subject to at least three limitations. First, during the review period some hospitalized patients might have had H1N1 but were not tested or confirmed, resulting in underreporting of cases. Second, despite the use of a standardized abstraction tool, incomplete information in the medical charts might have led to underreporting of some underlying illnesses and limited the ability to study their role in the development of severe influenza. Finally, patients hospitalized in the first weeks of the outbreak likely do not represent patients later hospitalized with H1N1 (>900 in New York City as of July 2009, after which surveillance was limited to sentinel hospitals and passive reporting).
Collecting data from the medical charts of hospitalized patients during the initial aspects of such epidemics can provide information useful to health departments for policy making or education or prevention campaigns, but the utility of such surveys must be balanced with the extensive resources required to collect such information. Currently, DOHMH is collecting clinical and laboratory data from patients with H1N1 infection at sentinel hospital sites. In particular, efforts are underway to collect height and weight to evaluate whether obesity is an independent risk factor for hospitalization. Public education campaigns should encourage patients at high risk of severe illness to be vaccinated, and should emphasize to medical providers the importance of early antiviral therapy for children aged <2 years and patients with underlying risk conditions.1
This report is based, in part, on contributions by JM Norton, PhD, and S Lim, MS, and the DOHMH 2009 Pandemic Influenza A (H1N1) Chart Abstraction Team, New York City Department of Health and Mental Hygiene.
In the early days of the 2009 pandemic influenza A (H1N1) outbreak, little was known regarding the risk factors or expected clinical course of H1N1 infection among hospitalized patients.
What is added by this report?
Detailed examination of hospitalization and patient outcome data during the initial outbreak of H1N1 influenza in New York City showed that 56% of adult patients hospitalized for H1N1 were obese, 92% of obese patients had other underlying medical conditions, and suggested that prompt antiviral therapy after symptom onset might be associated with shorter length of hospitalization.
What are the implications for public health practice?
This rapid assessment led to a greater understanding of the disease; results were used by New York City health authorities when issuing guidance to the public and providing information and health alerts to New York City health-care providers.
Patients Hospitalized With 2009 Pandemic Influenza A (H1N1)—New York City, May 2009. JAMA. 2010;303(7):604-607. doi: