Introduction

The emergence of the novel H1N1 influenza strain in 2009 garnered a great deal of global attention. The H1N1 virus which emerged as a novel virus was first detected in the United States in April 2009. This prompted the U.S. Department of Health and Human Services to declare a public health emergency on April 26, 2009; to better prepare for the effects this novel virus could cause [1]. As a result, public health agencies across the country increased surveillance efforts in an attempt to quickly contain the situation.

On the average, seasonal flu infects between 5 and 20 percent of people in the United States each year, causing approximately 36,000 deaths annually [2]. Although most cases were self-limiting, H1N1 influenza spread more rapidly than seasonal influenza in 2009 and fatalities occurred in many countries [3, 4]. Early reports of illnesses associated with novel influenza A (H1N1) infection indicated that the disease might be similar in severity to seasonal influenza, with majority of patients not requiring hospitalization and deaths rarely reported in persons with underlying medical conditions [5]. While everyone was at risk for contracting H1N1, in 2009 the number of cases and deaths reported from H1N1 were less among people over the age of 64 years than that of the seasonal flu [6]. In addition, the Centers for Disease Control and Prevention (CDC) reported that novel H1N1 flu has caused greater disease burden in people younger than 25 years of age than older people [7].

The advent of the novel H1N1 virus prompted HDHHS to use the existing sentinel surveillance system to effectively monitor the situation of novel H1N1 and other seasonal flu viruses in the Houston metropolitan area. Although there are many similarities between the H1N1 virus and the seasonal flu, they are two different strains of influenza. The occurrence of this new strain of influenza virus coupled with intense efforts to control it, offer a unique opportunity to document its key epidemiological properties. This paper therefore provides an epidemiological analysis and comparison of the cases of seasonal and novel H1N1 influenza reported to HDHHS by healthcare providers that formed the sentinel surveillance network in Houston, Texas during the period between October 2008 and October 2009.

Methods

During the influenza season of 2007–2008, the Houston Department of Health and Human Services (HDHHS) initiated a partnership with local providers including private clinics, schools, and public hospitals. The aim was to develop a sentinel surveillance system for near real-time monitoring of respiratory viruses circulating within the community. A total of 30 providers were randomly selected using the probability proportional to size (PPS) sampling technique from a pool of 310 facilities that were participating in the “Vaccine for Children” program of the HDHHS. The sampling process was based on compliance of providers within the “Vaccine for Children” program as well as the geographic distribution of the clinics in the city of Houston, Texas.

The providers were initially grouped into two main categories based on their compliance to 4-3-1-3-3 coverage rates [8]. These included the compliance group (above 80%) and non-compliance group (below 80%). The compliance group was further divided into four geographical quadrants namely Southwest (SW), Southeast (SE), Northwest (NW) and Northeast (NE). The sample of the compliant group was further weighted based on the number of providers in each quadrant of the city. About 76.7% of the sample size was selected from the compliant group, while the remaining 23.3% came from the non-compliance providers. The non-compliant group was added to get a representative sample of city of Houston providers.

The selected providers were contacted and formal partnerships were established. They were trained on various aspects of the surveillance techniques including specimen collection, storage, and transportation. HDHHS provided free rapid flu testing and specimen collection kits to the participating providers.

The providers collected specimens from all patients with ILI symptoms visiting their clinics and submitted these to HDHHS. The HDHHS laboratory then tested the specimens for the novel H1N1 and other seasonal flu viruses using real-time reverse transcriptase-polymerase chain reaction (RT–PCR), and results were reported to both the HDHHS influenza surveillance team and the providers. The CDC approved case criteria, definitions, procedures and testing for patients with suspected novel influenza A (H1N1) virus infection were used [9]. All influenza A specimens that were not subtypeable at HDHHS laboratory, were sent to the Viral Surveillance and Diagnostic Branch of the CDC’s Influenza Division as soon as possible for further diagnostic testing. All PCR-confirmed influenza cases were further investigated using medical chart abstractions and patient interviews by HDHHS epidemiologists. Information collected included demographic data and details of clinical illness (e.g. date of onset, signs and symptoms, severity, medications given, and outcome).

Household contacts exhibiting similar symptoms were also investigated. Four attempts were made to contact all patients. Patients who could not be contacted after the attempts were declared as lost-to-follow-up. All information obtained were entered into MAVEN, an integrated public health database system. Cases with inconclusive or unsatisfactory PCR results were not included in the data analysis, and those with missing information were also excluded from that particular analysis. Data obtained were subjected to descriptive and inferential statistical analyses. Differences for measure of independent association between variables were explored using chi-square test statistics. All statistical test were two-sided, and statistical significance was defined at P < 0.05. Data Analyses were conducted using SAS statistical software, version 9.1.3 (SAS Institute Inc, NC, USA) [10].

Results

Figure 1 presents the flowchart of ILI cases reported to the HDHHS and the laboratory outcome of test conducted on specimens. Overall a total of 1,122 ILI cases were reported to HDHHS by sentinel providers and tested by HDHHS laboratory. Of this number, 296 (67.5%) specimens tested positive for influenza A; 140 (32.0%) for influenza B, and 2 (0.46%) for influenza A/B. Two hundred and fifty-nine (59%) were confirmed cases of seasonal influenza and 179 (41%) were novel H1N1 subtype. The demographic distribution of the ILI cases reported to HDHHS by the sentinel providers are given in Table 1, with 58.4% of the cases reported being from children ages 1–9 years. Also, a male to female ratio of 1:1 was noted for all ILI cases.

Fig. 1
figure 1

Flowchart of ILI cases reported to HDHHS and the laboratory outcome of test conducted on specimens by sentinel providers

Full size image
Table 1 Distribution of ILI cases reported to HDHHS by sentinel providers between October 2008 and October 2009
Full size table

Table 2 presents the distribution of all positive flu cases. The results revealed that a significant difference (P < .0001) existed within the age groups, with majority of the patients being in the age group 5–9 years for both seasonal and novel H1N1 influenza. This was closely followed by the age category 1–4 years and 10–14 years for patients infected with seasonal influenza and novel H1N1 virus, respectively. The median ages for seasonal influenza and novel H1N1 were 7 and 8 years, and differed only slightly from the median age of 6 years recorded for all ILI cases. The individuals in age group 40 years and above accounted for the fewest number of cases for both seasonal influenza and novel H1N1. A fairly equal and non-significant gender distribution was noted. There was a significant racial variation (P ≤ 0.01) in the number of positive cases for both disease types. The majority of cases reported to HDHHS were of Hispanic ethnic background followed by Whites and African-Americans. However, when examining the demographic distribution between seasonal and novel H1N1 influenza, no significant differences were found for any of the variables considered (Table 2).

Table 2 Distribution of confirmed cases of seasonal and novel H1N1 influenza reported to HDHHS by demographic characteristics
Full size table

Figure 2 describes the seasonal trends of seasonal influenza and novel H1N1 cases reported to HDHHS during the study period. Seasonal influenza activities started in December 2008 with a peak in February 2009. The activities declined in the following months with a recurrence in September 2009. Novel H1N1 virus activity was initially reported in May 2009 with a peak in September 2009, the same month seasonal influenza made resurgence. After September, novel H1N1 cases declined.

Fig. 2
figure 2

Trends in confirmed cases of Seasonal and novel H1N1 Influenza reported to HDHHS during 2008–2009 flu season

Full size image

Findings from our data indicated that the most frequently reported symptom in both seasonal flu and novel H1N1 cases was fever (83.3% vs. 91.8%), followed by cough (30.0% vs. 44.3%). In addition, a fair amount of patients infected with novel H1N1 virus also reported having sore throat, rhinorrhea, and muscle ache (Table 3). Information on number of days ill was incomplete as only ninety-two of the positive cases had data available on number of days ill. The median time ill for seasonal influenza cases was 6 days (n = 63, range of 1–47 days) while median time ill for novel H1N1 cases was 2 days (n = 16, range of 1–6 days). It was not possible to obtain medical history and follow-up interviews for all influenza cases in this study population. A total of 65 seasonal influenza cases were interviewed and of those, 15 (23%) reported prior seasonal influenza vaccination. The median missed time of work or school for interviewed patients with available information was 3 days (n = 53; range: of 0–6 days).

Table 3 Symptoms of seasonal influenza and novel H1N1 influenza casesa
Full size table

Discussion

During major outbreaks, early and accurate diagnosis of influenza has the ability to ensure prompt and appropriate treatment and can ultimately decrease the economic and public health burden [11]. Since the emergence of H1N1 as a novel virus, the epidemiologic profile of the disease has varied considerably depending on the country and continent being examined [12]. The results from this study provide general information regarding demographic characteristics and symptomatic attributes of seasonal influenza and novel H1N1 cases in Houston, Texas. Our findings indicate that majority of the cases for seasonal influenza (62.6%) occurred in children ages 1–9 years in contrast to novel H1N1 virus, where 63% of the cases were recorded among children 5–14 years old. These findings support earlier reports from other researchers, where higher incidence rates were also observed among younger age groups [11, 13]. CDC report indicates that novel H1N1 flu was most prevalent among people younger than 25 years of age [14]. In Japan, its been reported that the number of hospitalized cases was highest among those aged 5–9 years, and very low among adult age groups [13]. The low number of cases reported among individuals of age 40 and above, in our study sample supports earlier findings that cross-protection may have occurred among older age groups owing to their prior exposure to H1N1 strains circulating before 1957, or minimal contact by older age groups with young travelers and school-aged children who amplify transmission during the early stage of the spread of a new influenza virus [5, 15]. Therefore, the demographic characteristics of cases documented in our study may not reflect the overall population at risk of infection in Houston, Texas, but rather the population contributing to the transmission dynamics described earlier [15]. Conversely, it may be associated with the selection of providers from the HDHHS “Vaccine for Children” program, which may not have an accurate representation of influenza infection in the elderly population in Houston, Texas.

There was no significant difference (P > 0.05) noted between the demographic characteristics of patients infected with seasonal and novel H1N1 virus in our sample, with reference to age, gender and race/ethnicity. Belongia et al. provided an excellent epidemiologic comparison of the clinical characteristics of the 2009 influenza A H1N1 versus other seasonal influenza A strains, [16]. They found that clinical manifestations and risk for hospitalization were similar between the 2009 H1N1 strain and other seasonal influenza A strains. However, children were disproportionately affected by the 2009 H1N1 strain but not necessarily by severity of illness [16].

Seasonal and novel H1N1 influenza were found to be fairly equally distributed between males and females. However, the low sample size did not permit stratified analysis to determine, if distribution of H1N1 differed by gender between age groups. Compared to other race/ethnicities, a greater proportion of Hispanics presented with ILI symptoms and also tested positive for both seasonal influenza and novel H1N1 virus in this study. Multiple factors may be associated with this disparity. The national Hispanic population’s median age (26 years) is much lower than that of non-Hispanics (37 years), [17] and younger age groups were found to have a higher prevalence of seasonal influenza and novel H1N1 in the current study. The Center for Disease Control and Prevention (CDC) reports that socioeconomic disparities and lower influenza vaccination rates among Hispanics may also contribute to disproportionally higher morbidity rates [18].

The time trend of influenza morbidity obtained from this analysis is consistent with that of other sources. The CDC recorded similar peaks of national influenza morbidity in the months of February, June, and September of 2009 [19]. Google Flu trends also indicated peaks in February and September of 2009 for the City of Houston, Texas [20]. These correlations indicate that data obtained through the HDHHS influenza sentinel surveillance system gave an accurate representation of influenza illness in Houston, Texas.

The symptoms of novel H1N1 virus seen among patients were similar to the symptoms of seasonal influenza although in addition, sore throat, rhinorrhea and muscle ache were more commonly associated with novel H1N1 virus than seasonal influenza. Fever and cough have generally been reported as the most common symptoms associated with novel H1N1 [3, 7]. Similarly, fever and cough have been reported to be the most highly predictive symptoms of seasonal influenza in adolescents and adults [21, 22] and cough (but not fever) in children 5–12 years [23]. However, we noted in the current study that fever was the most frequently reported symptom for both seasonal and novel H1N1 influenza (83.3% vs. 91.8%), followed by cough (30.0% vs. 44.3%). It has been argued that when influenza is known to be circulating in the community, these symptoms alone may be sufficient to diagnose influenza in otherwise healthy individuals, [21, 22] although this may not always be the case. We noted that 23% of the 65 patients interviewed indicated that they had prior seasonal influenza vaccination. Similarly, Hahné et al. documented that of the 111 cases of influenza A (H1N1) reported, 17 (15%, 95% CI 9–23%) of the patients had received vaccination for 2008–2009 flu season [24]. These observations are consistent with the fact that receipt of recent (2005–2009) seasonal influenza vaccines was unlikely to elicit a protective antibody response to the novel influenza A (H1N1) virus [25]. However, the lack of significant changes in the novel H1N1 virus may indicate that the vaccine being manufactured would closely match the circulating viruses and provide people with effective protection against novel H1N1 virus and other seasonal viruses.

The median days ill among our sample population was 6 days for seasonal influenza cases and 2 days for novel H1N1 cases. While a higher median of 4 days (range 0–23 days) have been reported for novel H1N1 virus in Ontario, Canada [26], our value falls within the range of 1–5 days given as estimate for novel H1N1 virus, and said to be shorter and more like seasonal influenza than originally thought [27]. The duration of illness typically reported for H1N1 in European countries ranged from 4 to 6 days with the infectious period for a confirmed case being 1 day prior to the onset of symptoms to 7 days after onset [15].

Limitations

This study was based on 30 participating providers, which represented only 10% of facilities involved in the “Vaccine for Children” program in Houston, Texas. Accordingly, data used for the current study may not be representative of the Houston population due to reporting bias, since only a fraction of persons with influenza symptoms. Are likely to have been tested? All diagnostic testings were clinically driven and therefore, the ability to detect cases depended on health seeking behaviors, testing behaviors, the resources available to health-care providers and the sensitivity and specificity of laboratory tests. Despite the use of standardized data collection form, not all information was collected for every patient. The small sample size coupled with low number of adult cases in the current study population was also considered a limitation. It is also important to consider biases in surveillance data, such as missing data and recall bias, when interpreting these results. The use of a larger patient sample that includes all age categories and spans more than one flu season may accurately represent the characteristics of the novel H1N1 virus in the community. Other areas that require further investigation include determining transmission dynamics, seroprevalence and risk factors for novel H1N1 infection and hospitalization.

Conclusions

Sentinel surveillance system offer advantages over passive surveillance in early identification of illness clusters, and could assist in mobilizing rapid response, thereby reducing morbidity and mortality. The findings presented in this paper highlight the comparative epidemiologic characteristics of both seasonal influenza and novel H1N1 virus in the Houston metropolitan area. The study further confirmed the similarity of the symptoms of novel H1N1 illness to that of normal seasonal influenza, and suggests that people with high risk for seasonal flu may also be at high risk for novel H1N1 infection. The sentinel surveillance system was helpful in providing HDHHS with timely information that assisted in monitoring the spread of both seasonal influenza and novel H1N1 virus and in executing early response activities in the Houston metropolitan area. However, continued and expanded surveillance is of key importance, and necessary to further characterize the pandemic and to support public-health planning.