Douglas M. Haas, MD

Study Design and Setting

This retrospective cohort study was conducted at a tertiary care center with 1159 beds divided between 2 general hospitals (634 and 167 beds), a large cancer center (208 beds), and an inpatient cardiac hospital (150 beds).

Selection of Participants

During the pandemic influenza season, clinicians throughout the tertiary care center were encouraged to obtain nasopharyngeal swabs to screen for influenza from all patients presenting to the hospital with respiratory illness, unexplained gastrointestinal symptoms, ILI, or any suspicion for influenza infection based on physician discretion. Patients were admitted through routine hospital admission procedures and triage. Physicians were encouraged to treat all patients in whom pandemic H1N1/09 was suspected empirically with an antiviral drug (oseltamivir). Patients with a positive H1N1/09 test were then continued on treatment. Obese patients, critically ill patients requiring intensive care unit (ICU) care, pregnant women, and patients with chronic medical conditions such as chronic respiratory disease, chronic kidney disease, diabetes mellitus, and immune‐compromising conditions were treated with a higher dose of oseltamivir (150 mg twice a day). Other patients were treated with the standard dose of oseltamivir (75 mg twice a day). The present analyses include all patients at least 18 years of age, hospitalized with laboratory‐confirmed pandemic H1N1/09 virus infection from 1 June 2009 through 31 December 2009. Subjects younger than 18 years and prisoners were excluded.

Data Collection

Laboratory‐confirmed pandemic H1N1/09 patients were identified from the hospital's infection control database. After identifying eligible patients, physician investigators accessed each electronic medical record to obtain health history, diagnosis, test results, disposition, and clinical outcome information. We captured data on gender, race, age, comorbidities, symptoms and onset of illness, hospital presentation, diagnosis, treatment, ICU admission, and death. Data abstraction was performed by an infectious disease physician using a standardized form, and all data abstraction was validated by another infectious disease physician. The research was deemed exempt from further review by the Ohio State University Institutional Review Board (2009E0979).

Subjects with laboratory‐confirmed H1N1/09 virus infection were identified if nasal swab or respiratory secretions were positive for novel influenza A (H1N1) by specific rapid antigen or culture testing. Nasal swabs that were processed for confirmatory influenza testing by polymerase chain reaction (PCR) were confirmed by panel (xTAG) or specific influenza A and B PCR (Luminex Molecular Diagnostics, Toronto, CA).14 Positive tests were confirmed using the Prodesse ProFlu‐ST, Influenza A [2009] real‐time PCR (H1N1 subtyping) assay (Focus Diagnostics, Cypress, CA).

Day of onset of illness was defined as the day when the patient started developing one or more symptoms of influenza‐like illness. Day of hospital presentation was defined as the day when the patient either presented to the emergency department or a physician's office and was subsequently admitted or was directly admitted to the hospital. Fever was defined as a temperature of 100F (37.8C) or greater at presentation, or reported subjective fever upon presentation, and reported antipyretic use (nonsteroidal anti‐inflammatory drugs or acetaminophen) prior to admission. When multiple temperature readings were present in the emergency department record, any recorded temperature of 100F or greater was considered a fever. The Charlson comorbidity index predicts the 10‐year mortality for a patient with a range of comorbid conditions such as heart disease, acquired immune deficiency syndrome (AIDS), or cancer. The Charlson comorbidity index was calculated according to previously described methods.1517 Diabetes, chronic respiratory disease (asthma, chronic obstructive pulmonary disease), and chronic kidney disease with hemodialysis were recorded as comorbidities if these diagnoses were documented in the patient notes. Immunosuppression was defined as oral steroid use or other immunosuppressive medication, organ transplantation, human immunodeficiency virus (HIV) infection, or cancer chemotherapy. Empiric antibiotic treatment was defined as treatment with antibiotics regardless of the evidence of bacterial infection. Proven bacterial coinfection was defined as infection with bacterial pathogens documented by a positive bacterial culture from any site that was obtained in response to clinical suspicion of bacterial infection.

Data Analysis

Demographic and clinical characteristics were summarized for all patients, and for those patients with or without fever, as part of their illness prior to hospital presentation. Continuous variables were summarized as mean ( standard deviation) or median values. For categorical variables, the percentage of patients in each category was calculated. For comparison between patients with fever and those with no fever, for continuous variables, the 2‐sample t test for normally distributed data or the Wilcoxon rank‐sum test for non‐normal data was used. The Pearson chi‐square test was used for categorical data. We performed a logistic regression analysis to examine the association between fever and time to presentation and ICU admission, adjusting for patient age and Charlson index. All analyses were carried out using SAS 9.2 software (SAS, Cary, NC).

Limitations

Our study took place in a single medical center, and our findings may not be generalizable to other clinical settings. However, this large tertiary care center admits 25% of patients in central Ohio. Our study focused on hospitalized patients only and excluded patients who may have presented to the emergency department with H1N1/09 but were not hospitalized. Thus, our findings can be applied to patients with severe enough disease to warrant hospitalization. It is possible that patients without fever presented early (within 2 days of onset of illness) and were treated as outpatients; these individuals would not have been captured in the current analysis. However, our primary goal was to assess the effect of fever on the overall course of H1N1/09 presentation, diagnosis, and management in hospitalized patients. Provider bias may have played a role in the decision to admit patients to the hospital, but clinicians were encouraged to use a standardized protocol to drive decisions to test and admit. We studied H1N1/09 patients only, and our findings may not be generalizable to patients infected with other influenza types. However, H1N1/09 influenza symptoms are similar to other types of influenza. We also did not collect data on insurance status which may play a role in patients' choice to seek hospital care.

Strengths

Our study also has substantial strengths. Because clinicians tested all patients with symptoms of respiratory illness and unexplained gastrointestinal symptoms irrespective of the presence of fever, we were able to assess the proportion of H1N1/09 patients without fever as part of their illness and to determine that the most common presenting symptoms were respiratory. Our analysis includes only patients with laboratory‐confirmed H1N1/09 infection. Our surveillance was strengthened by the availability of real‐time PCR confirmatory testing performed in our own molecular microbiology laboratory.

Conclusions

In this study, nearly half of patients with H1N1/09 requiring hospitalization did not have fever as part of their illness. Patients with fever tended to seek medical care sooner and thus received treatment sooner. Patients with late presentation and no fever were more likely to need ICU admission. Our findings reinforce previous studies that indicate better outcomes with early treatment of influenza.1013 These findings can be used to prompt clinicians to consider treating hospitalized patients with influenza and to educate the public to seek medical care soon after the onset of illness, despite absence of fever, if symptoms of respiratory illness and unexplained gastrointestinal symptoms are present during the influenza season.

Study Design and Setting

This retrospective cohort study was conducted at a tertiary care center with 1159 beds divided between 2 general hospitals (634 and 167 beds), a large cancer center (208 beds), and an inpatient cardiac hospital (150 beds).

Selection of Participants

During the pandemic influenza season, clinicians throughout the tertiary care center were encouraged to obtain nasopharyngeal swabs to screen for influenza from all patients presenting to the hospital with respiratory illness, unexplained gastrointestinal symptoms, ILI, or any suspicion for influenza infection based on physician discretion. Patients were admitted through routine hospital admission procedures and triage. Physicians were encouraged to treat all patients in whom pandemic H1N1/09 was suspected empirically with an antiviral drug (oseltamivir). Patients with a positive H1N1/09 test were then continued on treatment. Obese patients, critically ill patients requiring intensive care unit (ICU) care, pregnant women, and patients with chronic medical conditions such as chronic respiratory disease, chronic kidney disease, diabetes mellitus, and immune‐compromising conditions were treated with a higher dose of oseltamivir (150 mg twice a day). Other patients were treated with the standard dose of oseltamivir (75 mg twice a day). The present analyses include all patients at least 18 years of age, hospitalized with laboratory‐confirmed pandemic H1N1/09 virus infection from 1 June 2009 through 31 December 2009. Subjects younger than 18 years and prisoners were excluded.

Data Collection

Laboratory‐confirmed pandemic H1N1/09 patients were identified from the hospital's infection control database. After identifying eligible patients, physician investigators accessed each electronic medical record to obtain health history, diagnosis, test results, disposition, and clinical outcome information. We captured data on gender, race, age, comorbidities, symptoms and onset of illness, hospital presentation, diagnosis, treatment, ICU admission, and death. Data abstraction was performed by an infectious disease physician using a standardized form, and all data abstraction was validated by another infectious disease physician. The research was deemed exempt from further review by the Ohio State University Institutional Review Board (2009E0979).

Subjects with laboratory‐confirmed H1N1/09 virus infection were identified if nasal swab or respiratory secretions were positive for novel influenza A (H1N1) by specific rapid antigen or culture testing. Nasal swabs that were processed for confirmatory influenza testing by polymerase chain reaction (PCR) were confirmed by panel (xTAG) or specific influenza A and B PCR (Luminex Molecular Diagnostics, Toronto, CA).14 Positive tests were confirmed using the Prodesse ProFlu‐ST, Influenza A [2009] real‐time PCR (H1N1 subtyping) assay (Focus Diagnostics, Cypress, CA).

Day of onset of illness was defined as the day when the patient started developing one or more symptoms of influenza‐like illness. Day of hospital presentation was defined as the day when the patient either presented to the emergency department or a physician's office and was subsequently admitted or was directly admitted to the hospital. Fever was defined as a temperature of 100F (37.8C) or greater at presentation, or reported subjective fever upon presentation, and reported antipyretic use (nonsteroidal anti‐inflammatory drugs or acetaminophen) prior to admission. When multiple temperature readings were present in the emergency department record, any recorded temperature of 100F or greater was considered a fever. The Charlson comorbidity index predicts the 10‐year mortality for a patient with a range of comorbid conditions such as heart disease, acquired immune deficiency syndrome (AIDS), or cancer. The Charlson comorbidity index was calculated according to previously described methods.1517 Diabetes, chronic respiratory disease (asthma, chronic obstructive pulmonary disease), and chronic kidney disease with hemodialysis were recorded as comorbidities if these diagnoses were documented in the patient notes. Immunosuppression was defined as oral steroid use or other immunosuppressive medication, organ transplantation, human immunodeficiency virus (HIV) infection, or cancer chemotherapy. Empiric antibiotic treatment was defined as treatment with antibiotics regardless of the evidence of bacterial infection. Proven bacterial coinfection was defined as infection with bacterial pathogens documented by a positive bacterial culture from any site that was obtained in response to clinical suspicion of bacterial infection.

Data Analysis

Demographic and clinical characteristics were summarized for all patients, and for those patients with or without fever, as part of their illness prior to hospital presentation. Continuous variables were summarized as mean ( standard deviation) or median values. For categorical variables, the percentage of patients in each category was calculated. For comparison between patients with fever and those with no fever, for continuous variables, the 2‐sample t test for normally distributed data or the Wilcoxon rank‐sum test for non‐normal data was used. The Pearson chi‐square test was used for categorical data. We performed a logistic regression analysis to examine the association between fever and time to presentation and ICU admission, adjusting for patient age and Charlson index. All analyses were carried out using SAS 9.2 software (SAS, Cary, NC).

Limitations

Our study took place in a single medical center, and our findings may not be generalizable to other clinical settings. However, this large tertiary care center admits 25% of patients in central Ohio. Our study focused on hospitalized patients only and excluded patients who may have presented to the emergency department with H1N1/09 but were not hospitalized. Thus, our findings can be applied to patients with severe enough disease to warrant hospitalization. It is possible that patients without fever presented early (within 2 days of onset of illness) and were treated as outpatients; these individuals would not have been captured in the current analysis. However, our primary goal was to assess the effect of fever on the overall course of H1N1/09 presentation, diagnosis, and management in hospitalized patients. Provider bias may have played a role in the decision to admit patients to the hospital, but clinicians were encouraged to use a standardized protocol to drive decisions to test and admit. We studied H1N1/09 patients only, and our findings may not be generalizable to patients infected with other influenza types. However, H1N1/09 influenza symptoms are similar to other types of influenza. We also did not collect data on insurance status which may play a role in patients' choice to seek hospital care.

Strengths

Our study also has substantial strengths. Because clinicians tested all patients with symptoms of respiratory illness and unexplained gastrointestinal symptoms irrespective of the presence of fever, we were able to assess the proportion of H1N1/09 patients without fever as part of their illness and to determine that the most common presenting symptoms were respiratory. Our analysis includes only patients with laboratory‐confirmed H1N1/09 infection. Our surveillance was strengthened by the availability of real‐time PCR confirmatory testing performed in our own molecular microbiology laboratory.

Conclusions

In this study, nearly half of patients with H1N1/09 requiring hospitalization did not have fever as part of their illness. Patients with fever tended to seek medical care sooner and thus received treatment sooner. Patients with late presentation and no fever were more likely to need ICU admission. Our findings reinforce previous studies that indicate better outcomes with early treatment of influenza.1013 These findings can be used to prompt clinicians to consider treating hospitalized patients with influenza and to educate the public to seek medical care soon after the onset of illness, despite absence of fever, if symptoms of respiratory illness and unexplained gastrointestinal symptoms are present during the influenza season.

Study Design and Setting

This retrospective cohort study was conducted at a tertiary care center with 1159 beds divided between 2 general hospitals (634 and 167 beds), a large cancer center (208 beds), and an inpatient cardiac hospital (150 beds).

Selection of Participants

During the pandemic influenza season, clinicians throughout the tertiary care center were encouraged to obtain nasopharyngeal swabs to screen for influenza from all patients presenting to the hospital with respiratory illness, unexplained gastrointestinal symptoms, ILI, or any suspicion for influenza infection based on physician discretion. Patients were admitted through routine hospital admission procedures and triage. Physicians were encouraged to treat all patients in whom pandemic H1N1/09 was suspected empirically with an antiviral drug (oseltamivir). Patients with a positive H1N1/09 test were then continued on treatment. Obese patients, critically ill patients requiring intensive care unit (ICU) care, pregnant women, and patients with chronic medical conditions such as chronic respiratory disease, chronic kidney disease, diabetes mellitus, and immune‐compromising conditions were treated with a higher dose of oseltamivir (150 mg twice a day). Other patients were treated with the standard dose of oseltamivir (75 mg twice a day). The present analyses include all patients at least 18 years of age, hospitalized with laboratory‐confirmed pandemic H1N1/09 virus infection from 1 June 2009 through 31 December 2009. Subjects younger than 18 years and prisoners were excluded.

Data Collection

Laboratory‐confirmed pandemic H1N1/09 patients were identified from the hospital's infection control database. After identifying eligible patients, physician investigators accessed each electronic medical record to obtain health history, diagnosis, test results, disposition, and clinical outcome information. We captured data on gender, race, age, comorbidities, symptoms and onset of illness, hospital presentation, diagnosis, treatment, ICU admission, and death. Data abstraction was performed by an infectious disease physician using a standardized form, and all data abstraction was validated by another infectious disease physician. The research was deemed exempt from further review by the Ohio State University Institutional Review Board (2009E0979).

Subjects with laboratory‐confirmed H1N1/09 virus infection were identified if nasal swab or respiratory secretions were positive for novel influenza A (H1N1) by specific rapid antigen or culture testing. Nasal swabs that were processed for confirmatory influenza testing by polymerase chain reaction (PCR) were confirmed by panel (xTAG) or specific influenza A and B PCR (Luminex Molecular Diagnostics, Toronto, CA).14 Positive tests were confirmed using the Prodesse ProFlu‐ST, Influenza A [2009] real‐time PCR (H1N1 subtyping) assay (Focus Diagnostics, Cypress, CA).

Day of onset of illness was defined as the day when the patient started developing one or more symptoms of influenza‐like illness. Day of hospital presentation was defined as the day when the patient either presented to the emergency department or a physician's office and was subsequently admitted or was directly admitted to the hospital. Fever was defined as a temperature of 100F (37.8C) or greater at presentation, or reported subjective fever upon presentation, and reported antipyretic use (nonsteroidal anti‐inflammatory drugs or acetaminophen) prior to admission. When multiple temperature readings were present in the emergency department record, any recorded temperature of 100F or greater was considered a fever. The Charlson comorbidity index predicts the 10‐year mortality for a patient with a range of comorbid conditions such as heart disease, acquired immune deficiency syndrome (AIDS), or cancer. The Charlson comorbidity index was calculated according to previously described methods.1517 Diabetes, chronic respiratory disease (asthma, chronic obstructive pulmonary disease), and chronic kidney disease with hemodialysis were recorded as comorbidities if these diagnoses were documented in the patient notes. Immunosuppression was defined as oral steroid use or other immunosuppressive medication, organ transplantation, human immunodeficiency virus (HIV) infection, or cancer chemotherapy. Empiric antibiotic treatment was defined as treatment with antibiotics regardless of the evidence of bacterial infection. Proven bacterial coinfection was defined as infection with bacterial pathogens documented by a positive bacterial culture from any site that was obtained in response to clinical suspicion of bacterial infection.

Data Analysis

Demographic and clinical characteristics were summarized for all patients, and for those patients with or without fever, as part of their illness prior to hospital presentation. Continuous variables were summarized as mean ( standard deviation) or median values. For categorical variables, the percentage of patients in each category was calculated. For comparison between patients with fever and those with no fever, for continuous variables, the 2‐sample t test for normally distributed data or the Wilcoxon rank‐sum test for non‐normal data was used. The Pearson chi‐square test was used for categorical data. We performed a logistic regression analysis to examine the association between fever and time to presentation and ICU admission, adjusting for patient age and Charlson index. All analyses were carried out using SAS 9.2 software (SAS, Cary, NC).

Limitations

Our study took place in a single medical center, and our findings may not be generalizable to other clinical settings. However, this large tertiary care center admits 25% of patients in central Ohio. Our study focused on hospitalized patients only and excluded patients who may have presented to the emergency department with H1N1/09 but were not hospitalized. Thus, our findings can be applied to patients with severe enough disease to warrant hospitalization. It is possible that patients without fever presented early (within 2 days of onset of illness) and were treated as outpatients; these individuals would not have been captured in the current analysis. However, our primary goal was to assess the effect of fever on the overall course of H1N1/09 presentation, diagnosis, and management in hospitalized patients. Provider bias may have played a role in the decision to admit patients to the hospital, but clinicians were encouraged to use a standardized protocol to drive decisions to test and admit. We studied H1N1/09 patients only, and our findings may not be generalizable to patients infected with other influenza types. However, H1N1/09 influenza symptoms are similar to other types of influenza. We also did not collect data on insurance status which may play a role in patients' choice to seek hospital care.

Strengths

Our study also has substantial strengths. Because clinicians tested all patients with symptoms of respiratory illness and unexplained gastrointestinal symptoms irrespective of the presence of fever, we were able to assess the proportion of H1N1/09 patients without fever as part of their illness and to determine that the most common presenting symptoms were respiratory. Our analysis includes only patients with laboratory‐confirmed H1N1/09 infection. Our surveillance was strengthened by the availability of real‐time PCR confirmatory testing performed in our own molecular microbiology laboratory.

Conclusions

In this study, nearly half of patients with H1N1/09 requiring hospitalization did not have fever as part of their illness. Patients with fever tended to seek medical care sooner and thus received treatment sooner. Patients with late presentation and no fever were more likely to need ICU admission. Our findings reinforce previous studies that indicate better outcomes with early treatment of influenza.1013 These findings can be used to prompt clinicians to consider treating hospitalized patients with influenza and to educate the public to seek medical care soon after the onset of illness, despite absence of fever, if symptoms of respiratory illness and unexplained gastrointestinal symptoms are present during the influenza season.