Introduction

Infection due to 2009 novel swine origin Influenza-A H1N1 virus (SOIV) has emerged as a global pandemic ⁽¹⁾. In the United States (US) and worldwide, several case series of severe disease and deaths have been reported. ^(2-9). In the US, New York City (NYC) was the first epicenter of the SOIV illness outbreak ⁽⁴⁾. US surveillance data reveals that 6.9% of all deaths were due to influenza and pneumonia, higher than the epidemic threshold of 6.6% ⁽¹⁰⁾. Currently, there are no studies on differences among characteristics of hospitalized SOIV infected inner-city minority adult and pediatric patients, except for a recent report on SOIV features among New Yorkers ⁽⁴⁾. Lincoln Medical and Mental Health center (LMMHC), a university-affiliated urban public NYC Health and Hospital Corporation institution serving inner-city minority New Yorkers in the South Bronx, encountered a number of cases with Influenza-Like Illness (ILI) during the 2009 outbreak. The characteristics and comparison of hospitalized pediatric and adult with SOIV illness are reported.

Methods and Materials

A retrospective observational study was conducted on adult and pediatric patients hospitalized for at least 24 hours with ILI and laboratory evidence of “probable” or “confirmed” 2009 SOIV-illness to LMMHC between April-July 2009. The study protocol and data abstraction tools were approved by the Institutional Review Board and informed consent was waived.

Patients with ILI were identified through daily reports from the emergency department(ED) and hospital wards. Data on demographic features, co-morbidity, clinical characteristics, laboratory tests and outcomes were extracted by 2 trained independent investigators using a standardized data-collection tool. Data regarding ED visits including respiratory infections utilizing ICD-9 codes were also collected.

Children were admitted to either pediatric ward or 8-bed pediatric intensive care unit (PICU) staffed 24/7 by on-call American board certified pediatricians and intensivists. Adult critically-ill patients were admitted to a closed medical ICU (MICU) staffed 24/7 by board certified critical care specialists. Standard ICU protocols were utilized in management of all patients. In medical ICU patients, length of stay, organ dysfunction, mechanical ventilation, Acute Physiology and Chronic Health Evaluation (APACHE) -II scores, and Sequential Organ Failure Assessment (SOFA) score on admission were noted. Organ dysfunction was defined using variables from the SOFA-scoring system with 0 being the best possible and 24 being the worst ⁽¹¹⁾.

Microbiological specimens were obtained using either a naso-pharyngeal swab or bronchial aspirate after endotracheal intubation in mechanically ventilated patients. Testing for influenza viruses was performed by a RIDT that detects influenza viral nucleoprotein antigen. This assay distinguishes influenza A/B viruses but not between influenza A subtypes. Positive RIDT results and selected patients with ILI with high suspicion and negative RIDT results were tested for influenza virus subtypes subsequently using Reverse Transcriptase- Polymerase chain reaction (RT-PCR) performed at the NYC Department of Health Public Health Laboratory.

ILI was defined as fever (temperature of 100°F [37.8°C] or greater) and cough and/or sore throat in the absence of a known cause other than influenza. A “confirmed” case of SOIV-illness was defined as a patient with an ILI and laboratory confirmed SOIV illness by RT-PCR. A “probable” case of SOIV illness is defined as a person with ILI, positive rapid antigen test for influenza-A, but did not have or tested negative by RT-PCR. Obesity was defined as a body mass index (BMI) of > 30 kg/m², or a BMI percentile of =/> 95 in children between the ages of 2 and 18 years, while overweight adults included those with BMI between 25-29 kg/m².

Statistical Methods

Descriptive data for adult and pediatric patients are presented separately and compared. Statistical analysis was performed using the Fischer’s exact test for categorical variables and Mann Whitney U test for continuous variables. A p value of < 0.05 was considered significant. Data was analyzed using STATA-version 8.0. The investigator analyzing the data was blinded to medical records.

Results

From April-July 2009, there were 55862 ED visits at LMMHC, with more than 16000 visits in May 2009 and 14000 in June 2009. About 12000 to 13000 patients were treated in the LMMHC ED every month in the last 2-year-period. The average number of ED visits during the pandemic increased to 13965 per month compared to an average of 12464 per month within the past 2 years, an increase of 12% every month (Figure-1).

A total of 2072 ED visits were related to a respiratory infection based on ICD-9 codes, which is a 25% increase in average number of visits related to respiratory infection during the 4-month study period (518 versus 416 visits). 1884 patients had ILI and underwent a RIDT. Of these, 298 were RIDT positive of which 80 patients including 40 adults and 40 children were hospitalized with confirmed or probable SOIV-illness. (Figure 2) Of the 40 adults, 35 were confirmed and 5 probable cases of SOIV-illness. In children, there were 39 confirmed and 1 probable cases. Among all, 14 tested RIDT positive, 5 had RT-PCR positive and 61 had both tests positive (Table-1).

SOIV-illness in minority adults

SOIV-illness in minority children

Comparing adults and children with SOIV-illness

Compared to children, adults had higher proportion of subjects with co-morbidity (75% vs.42.5%, p=0.039); organ failure (42.5% vs.0%, p-value=0.000); mechanical ventilation support (32.5% vs. 0%, p-value =0.000); median LOS (4.5 vs.3 days, p=0.0011) and deaths (10% vs.0%, p-value = 0.116).

Figure 1

Figure 1: LMMHC ED visits from Sept 07 to Aug 09

Figure 2

Figure 2: Patient selection

Figure 3

Table 1: SOIV illness test results

Figure 4

Table 2a: Clinical Characteristics of adults with probable or confirmed SOIV Illness

Figure 5

Table-2b: Clinical features Of Adult ICU Patients compared to Adult non-ICU patients with SOIV Illness

Figure 6

Table 3a: Demographic and Clinical features of children with probable or confirmed SOIV Illness

Figure 7

**Table 3b: Selected clinical characteristics across pediatric age groups

Discussion

This is the first study to describe separately and compare hospitalized adult and pediatric subjects during the 2009 SOIV-illness outbreak among urban inner-city residents of NYC. The ethnic distribution of our study subjects followed the population distribution of predominantly Hispanic and African-Americans in the South-Bronx ⁽¹²⁾. All of our study subjects were from the South Bronx zip-codes constituting the primary and secondary services areas of LMMHC. All children and 95% of adults had health insurance. Lack of insurance and access to care therefore does not seem to be a barrier in this study population. Adults at our institution had higher morbidity and mortality than children, a finding not seen in other reports ^(2-9).

Compared to seasonal-influenza¹³, adults hospitalized at our institution with SOIV illness were of younger age (median: 48 years), frequently had chronic illness (72.5%), and often had severe illness that required critical care, similar to other recent reports ^(7-10). Gastrointestinal (GI) symptoms of vomiting, abdominal pain and diarrhea were noted at presentation among a higher proportion of adults (32.5%). 2/3^rd of our adult patients were overweight/obese (65%) which is not a known risk factor for seasonal influenza. A higher but statistically non-significant proportion of obese individuals versus non-obese (29.7% vs.12.7%, p=NS) and overweight/obese combined versus non-obese/overweight adults (38.5% vs. 21.4%, p=NS), required invasive or non-invasive ventilation support, concurring with recent reports of severe illness among previously healthy obese subjects ⁽¹⁵⁾. Obese patients are likely to have decreased respiratory reserve or pre-existing lung conditions like obesity-hypoventilation syndrome. It is postulated that obesity may lead to altered immune function that leads to a decreased response to influenza viruses ⁽¹⁴⁾, although it is unclear why the 2009 SOIV afflicts obese subjects more. These factors combined with severe respiratory illness may explain the requirement for mechanical ventilation and MICU care among our SOIV-infected minority adult patients.

Our adult-MICU patients had relatively severe illness (median-SOFA score on admission 13 and APACHE-II score=31) compared with other reports on utilization of critical care resources for SOIV-illness ^(7-10). Many of them had severe-sepsis and required mechanical ventilation. Adult-MICU patients (53%) compared to non-ICU adult patients (16%) had a higher proportion of diabetes, but not obesity. Thrombocytopenia, abnormal CPK, and LDH elevations were noted frequently in MICU cases. Lymphopenia was present in 3/4^th of adult patients (<1000 cells/mm³) as reported in recent studies in the US and Mexico ^{(6, 8, 9).} Although adult-MICU cases had significantly more abnormal chest X-rays than non-MICU cases, 4 of the 15 MICU cases had normal chest-X ray on presentation. A unique feature noted among critically-ill adults was the poor reliability of RIDTs with naso-pharyngeal swabs (3/15 positive) compared to lower respiratory-tract specimens (12/15 positive). 2/3^rd of adults including all MICU-patients received anti-bacterial antibiotics at admission whereas only 2 patients had a confirmed isolate on cultures. Oseltamivir (Tamiflu) at standard recommended dose and frequency was initiated in all except 1 adult on the day of admission ⁽¹⁶⁾. 55% of adults and 30% of children presented with abnormal chest radiographs. Focal consolidation rather than an interstitial pattern of involvement in seasonal influenza was the most common type of radiological involvement in SOIV-illness. These findings could represent secondary or superimposed bacterial pneumonia that may be indistinguishable from the primary viral pneumonia previously described with the seasonal influenza ^{(17, 18)}. Corticosteroids were administered to patients with obstructive lung disease exacerbations (25%) and those with critical illness related corticosteroid insufficiency (CIRCI, 22.5%). The overall mortality rate among hospitalized adults was 10%. Median age of adults who died was 46.5 years (range 30-55years), many of them were obese, a finding not seen in seasonal influenza. The observed mortality among adult-ICU patients was 26 %, similar to reports from Mexico and Canada ^{(8, 9, and 19).}

Among children, GI symptoms were seen in a similar proportion (30%) as in patients with seasonal influenza, but less than among children in a national cohort of SOIV illness (42%). ⁽⁶⁾ Wheezing occurred in 3/10 asthmatic children and only in 4 patients without asthma. Also, one-quarter of children > 2 years of age were obese, comparatively less than in pediatric patients from the national US cohort (30%). Lymphopenia was common and noted in approximately half of our children (mostly among those older than 4 years of age) compared to other reports ⁽²⁰⁾. None of them had clinical myositis ⁽²¹⁾. No child had invasive bacterial co-infection, but 2/3^rd of cases received systemic antibiotics. Although bacteriology cultures were less revealing, focal radiograph findings in our study and 2 recent reports on H1N1 illness demonstrating frequent concurrent bacterial infections in 29-43% of subjects underscore the importance of early recognition and treatment of bacterial infection on admission in SOIV illness ⁽²²⁾. Abnormal chest radiographs were least common among youngest patients, while ½ of children > 5 years had abnormal chest x-rays, consolidation being the most common pattern as in adults. All children received Oseltamivir on day of admission and 9 pediatric patients received steroid therapy. Although 1/4^th of children were treated at the pediatric-ICU under close monitoring, none had organ dysfunction, or required ventilation support. We did not observe any deaths among pediatric subjects and this is contrary to reports of significant case fatality among children ⁽²³⁾. This can be related to several reasons. First, none of them had neuro-muscular disorder or immune suppression, the 2 factors associated with poor outcome in a recent report of pediatric deaths from SOIV illness ⁽²³⁾. Secondly, none of them had invasive bacterial co-infection as reported in a recent autopsy study on mortality related to SOIV-illness ⁽²³⁾.

According to CDC estimates, the baseline percentage of SOIV illness-related visits in the New York State region is 2.9% of all visits in 2009 ⁽²⁴⁾. Based on the volume of patient visits at LMMHC-ED, the expected number of influenza case visits is 287 cases per month. During the 4 month period, compared to the 1148 expected cases based on CDC estimates, more (1884) patients with ILI were seen in our ED. The 12% rise in total ED visits during this period suggests that there could have been more non-tested cases of SOIV illness.

The main limitations of our study are the relatively limited sample size and lack of non-minority controls. However, our data represents the largest sample that any single center has reported in the literature on this new emerging infectious disease, especially among urban inner city populations. Despite lack of non-minority controls, we compared our data with current reports of H1N1 cases in other US and international populations ^(2-9) and also provided age-based control comparisons within our sample.

Our study highlights the unique risk factors, age-related and ethnic disparities related to SOIV-illness among inner-city minority New Yorkers. Given the characteristics of the minority population in inner-city areas with increased prevalence of co-morbid conditions including diabetes, obesity and asthma, compounded by the decreased access to medical care that may result in delayed presentation, population mixing and overcrowding, decreased vaccination rates, the risk for adverse outcomes due to SOIV illness in these populations may be greater ^{(25, 26)}. In a busy urban inner-city hospital emergency room setting, patients at highest risk for severe SOIV-illness, should be promptly identified during triage and treatment established. Health care institutions serving urban inner-city populations may experience greater demand during pandemics related to excess volume of ED visits and increased need for hospitalization or critical care. Public health policy must consider and address the urban minority population risks and health care system preparedness in the national response to emerging pandemics and allocate appropriate resources.

Conclusion

In summary, SOIV-illness presents with unique clinical features among urban inner-city minority New-Yorkers. Our study found that morbidity and mortality due to SOIV-illness were greater among adults compared to children and associated with non-traditional risk factors such as obesity. Severity of illness and deaths were greater in young adults. The observed clinical features, morbidity and mortality among urban minority adult patients during the SOIV-pandemic are different from seasonal-influenza. A full understanding of this new emerging infection is essential to target those at risk for adverse outcomes, re-direct public health strategies and re-allocate resources to heath care institutions serving inner-city minority populations.

Acknowledgements

• Raghu Loganathan MD FCCP : Writing research protocol, creating data abstraction tool, writing IRB protocol and obtaining approval, data abstraction and review, manuscript preparation and review

• Yekaterina Sitnitskaya MD: Writing research protocol, creating data abstraction tool, presenting to IRB, data abstraction and review, manuscript preparation and review

• Melissa Schori MD MBA : Conceptualization of study, Data collection, Manuscript preparation and review

• Aliya Saeed MD: Data collection.

• Felix Rivera MD: Manuscript preparation and review.

• Jorge Bastos MD: Data collection.

• Yuly Carreras MD: Data Collection.

• Balavenkatesh Kanna MD, MPH, FACP : Writing the research protocol, Data compiling and review, manuscript preparation and review, statistical analysis