Prediction and surveillance of influenza epidemics

The 2009 (H1N1) influenza pandemic highlighted the need for forecasting models to assist policy and planning for effective health system responses to epidemics. Adequate forewarning of impending influenza epidemics would allow hospitals to adjust staff rosters, assign dedicated wards for infection control, and establish separate influenza clinics collocated with emergency departments (EDs).

We describe the use of surveillance and forecasting to predict and track influenza outbreaks. Although this work was instigated in response to the 2009 (H1N1) influenza outbreak, the methods apply to future outbreaks of other infectious diseases. There is extensive related research in this area using temporal, spatial, or space–time detection algorithms (eg, regression, smoothing, hidden Markov, and wavelet models), with successful performance reported for both sudden and gradual outbreaks.1-9 The objective of much of this modelling is to complement traditional public health monitoring with routine automated data analysis.1 Characteristics of outbreaks that have influenced detection include the magnitude of the signal, the shape of the signal, and the timing of the outbreak.2 An interesting observation is that, since 2000, prospective health surveillance has moved from weekly or monthly data, to daily and faster data rates.1 In general, for detecting changes in recent data behaviour, methods incorporating recent data history have advantages over those producing estimates from longer baselines.3

Emergency department data have frequently been used to identify disease patterns,10-15 including assessment of the timing and magnitude of daily influenza counts against laboratory confirmed influenza.10 We describe three approaches to predicting future epidemics based on emergency department data: surveillance monitoring of influenza presentations using adaptive cumulative sum (CUSUM) plan analysis to signal unusual activity; generating forecasts of expected numbers of presentations for influenza, based on historical data; and using internet search data as outbreak notification among a population. All three are considered applicable to health facilities that routinely collect and code patient arrival data.

We collected 5 years of historical data (2005–2009) on presentations and hospital admissions for influenza-like illnesses from the Emergency Department Information System (EDIS) database of 27 Queensland public hospitals. The influenza season assessed each year comprised June, July, August and September. Waiting times and triage categories are routinely reported for these 27 public hospitals.

Patients with influenza-like symptoms were identified by the following International Classification of Diseases (ICD-10-AM) codes within EDIS:

A08.4 – Viral gastroenteritis
B34.9 – Viral infection
J10.8 – (Influenza old code)
J11.1 – (Influenza new code)
J11.1S – H1N1 influenza (Human swine influenza) suspected
J18.0 – Bronchopneumonia
J18.1 – Lobar pneumonia unspecified
J18.2 – Hypostatic pneumonia unspecified
J18.8 – Pneumonia, atypical
J18.9 – Pneumonia, unspecified
Z04.8 – Medical review

Automatic confirmatory swab analysis for suspected influenza is not routine practice in Queensland EDs; hence, the inclusion of codes for possible cases, based on clinician experience. The specified symptom groups included in the codes listed were chosen in consultation with clinicians working at some of the participating hospitals. Although the codes included cover a broad set of conditions (low specificity), they are adequate to reliably capture the range of diagnoses of patients presenting with influenza.

Admitted patients (patients who required subsequent admission to a hospital bed) were defined using EDIS departure status description “admitted” (including to an ED bed), and “admitted to the ED” (ie, to a short-stay or observation ward).

With routinely collected health data becoming ubiquitous and relatively inexpensive to obtain, the potential for utilising forecasting and surveillance to support the decisions of hospital policymakers is evident.

Our CUSUM analysis identified an unusual influenza outbreak in the Australian summer of 2008–2009, as well as significantly more influenza-like presentations in the following winter. The latter outbreaks coincided with the worldwide H1N1 influenza outbreak. The results of the surveillance monitoring indicated that all hospitals, apart from one, had bigger than expected influenza outbreaks starting in late 2008 and continuing well into 2009. The one hospital that was not as affected by swine influenza was a large children’s hospital, which, interestingly, had relatively larger influenza outbreaks in 2007 and 2008 than other hospitals. It could be hypothesised that the high exposure of the source population to the 2007 and 2008 influenza made it less susceptible to influenza in 2009.

Many hospitals have seen the merit of generating forecasts from historical data to manage elective and emergency admissions in a proactive manner.17-22 The ability to accurately predict patient flow on a weekly or seasonal basis could enable more informed decisions about staffing levels to be made, which, in turn, could result in cost savings and improved patient care. Forecasting is a widely applicable, multidisciplinary science, encompassing statisticians, economists and operations researchers, that guides decision making in many areas of economic, industrial and scientific planning, but has gained little traction in the health care industry.22 We advocate that all health facilities that routinely collect data relating to workload and bed demand use this information proactively for improving management and patient outcomes. Our analysis shows, however, that routine forecasting methods need to be coupled with an adaptive methodology, such as the CUSUM approach described here, that can detect unusual occurrences.

The value of using internet search data in pandemic planning has been reported previously,23-25 with Google publishing correlations between search data and epidemics.23 Now these search data are accessible to the lay community and were used to construct the plots in Box 4. Additionally, Google’s Insights for Search site can report search interest relating to other infectious diseases to allow state health departments to survey trends for diseases other than influenza. According to Google, raw search volume is not reported, as the size of some regions may skew the data, and normalised data are used instead. The coherence of geographically specific search data with influenza presentations may enable information-poor regions to track the progression of pandemics in the community.

To determine unusual events and predict expected presentation rates, the best system among those assessed was a combination of routine forecasting methods coupled with an adaptive CUSUM methodology. Widely accessible internet search data can also assist by reflecting the timing of epidemics. The use of these forewarning and forecasting models can assist clinicians and hospital managers in clinical and infection control decision making.

1 Presentations (A) and admissions (B) of patients with influenza-like symptoms to public hospitals in Queensland in the influenza seasons of 2005–2009; the 2009 influenza season differed significantly from those of preceding years

2 CUSUM plans for signalling unusual presentations and admissions for influenza at three Queensland hospitals

There are 10 non-consecutive signals above the red line, indicating 10 different periods where counts were higher than expected given the influenza counts over the previous 6 days. As we expect only two false alarms for the period, most of these seem to be real alarms, indicating significantly higher than expected counts. It is fairly clear from the control chart above that 2009 was going to be a bad year, with several repeat signals in 2009 prior to the usual winter influenza season.

There are 12 different periods above the red line where counts were higher than expected, indicating late in 2008 and early in 2009 that 2009 was going to be a bad year for influenza. The control chart also identified when the outbreak stopped being unusual, with no alarms in August and September — the more traditional influenza season.

There are 15 non-consecutive signals above the red line, indicating significantly higher than expected counts. Note that the influenza outbreak in 2008 was not higher than expected, but from December 2008 alarms indicated significantly higher than expected counts despite the model adjusting for these. Either the transmission rates were higher than in previous years or the pathogen activity was higher. Again, it is clear that 2009 was going to be a bad year for influenza, and the control chart also identified when the outbreak stopped being unusual.

3 Prospective forecast accuracy for statewide data from 27 Queensland public hospitals, 2005–2009. Upper section (grey): all patients presenting to emergency departments (EDs). Lower section (green): patients with influenza-like diagnoses

Vertical axes: MAPE = Mean absolute percentage error (%) between forecasts and observed ED presentations. Horizontal axes: Month of the year denoted by numbers 1 to 12.

4 Correlation of statewide emergency department presentations for influenza and statewide internet searches related to influenza, Queensland

Vertical axes reflect data normalised in proportion to maximum data values.