The ‘Epidemiological Signal Detection’ Package

1. Datasets to be used in the Epidemiological Signal Detection Tool

Three types of datasets can be used in the tool:

• The default dataset included in the EpiSignalDetection package: EpiSignalDetection::SignalData;
• A basic ECDC Atlas export (csv format) exported from http://atlas.ecdc.europa.eu/public/index.aspx;
• Any dataset specified as described below.

#> Warning: `group_by_()` was deprecated in dplyr 0.7.0.
#> ℹ Please use `group_by()` instead.
#> ℹ See vignette('programming') for more help
#> Call `lifecycle::last_lifecycle_warnings()` to see where this warning was
#> generated.

2. Detection algorithms available in the tool

Two detection algorithms are available in the tool: the Farrington Flexible and the Generalized Likelihood Ratio statistic for Negative Binomial distribution (GLRNB). These algorithms are implemented in the R package surveillance ¹.

The Farrington Flexible algorithm takes range values of the surveillance time series and for each time point uses a Poisson GLM with over-dispersion to compute a predictive distribution for the current number of cases in case there is no outbreak - see Farrington et al. ² and Noufaily et al. ³ for details. A quantile, say the 99% quantile, of this predictive distribution then serves as threshold to determine if the actual observation is unusually high: If the observation is above the boundary, then an signal is raised. The Farrington method is a robust method that can detect the emergence of very rare serotypes with few cases as well as outbreaks for more common serotypes. It also takes seasonality and trend into account, and down-weights past unusually high number of cases. It has been widely used in by public health institutes ⁴.

This GLRNB algorithm is a count data regression chart for the monitoring of surveillance time series. It implements the seasonal count data chart based on generalized likelihood ratio (GLR) as described in the Hohle and Paul (2008) paper. A moving-window generalized likelihood ratio detector is used. Like the standard CUSUM method, it is sequential by nature. Unlike the CUSUM it does not require the specification of the alternative out-of-control hypothesis, but tries to estimate this in each time step. When overdispersion is present in the data, the dispersion parameter α is estimated as part of the in-control model estimation. However, it is estimated only once from the first fit.

The specific values of the parameters used for each algorithm function are set according to the parameter table included in the package:

my_parameters <- EpiSignalDetection::AlgoParam
names(my_parameters)
#> [1] "FarringtonFlexible" "GLRNB"

3. Interactive interface

The selection fields for parameters are located on the left-hand side of the application and are organised hierarchically (see Fig.1).

Each parameter has a default value. However, values available in the dropdown menus are hierarchically updated according to the previous parameter selection (see workflow described in Fig.2 below).

Fig.2 Workflow of panel update

In all dropdown boxes, the user can also choose to type in the values themselves - or start typing and select from options, rather than select from the dropdown menu (see Fig.3).

Fig.3 Dropdown boxes

In the first panel Dataset selection, user need to specify on which dataset to work on. The dropdown menus will display all values available in the given dataset:

• Health topic, disease name.
• Region name, geographical area. Please note that the default value “EU-EEA - complete series” is an additional geographical level from those available in the dataset. By selecting this option, the user will get the number of cases aggregated at EU-EEA level, excluding countries with gaps in the time series.
• Indicator, please note that only case-based indicators are allowed in the tool.
• Population.

The Time settings panel includes the selection of:

• Time unit, please note that the tool handles weekly and monthly data. For weekly aggregation, we use ISO8601-week.
• Study period. The default value is set to the most recent 5 years data available in the dataset. The minimum and the maximum are set to the full range of the time variable available in the dataset. The user can either type each date or use the calendar to define the start and the end of the study period (see Fig.4).

Fig.4 Calendar parameter

• Signal detection period. The default value of the signal detection period is:
  • 3 months out of 12 if time unit is “Month”
  • 13 weeks out of 53 if time unit is “Week”.

Fig.5 Signal detection period

The Signal detection algorithm panel allows to choose between the algorithms from the surveillance package available in the tool:

• Statistical algorithm using the parameters as defined in EpiSignalDetection::AlgoParam (see Tab.3 and Tab.4 above).

4. ‘Run the analysis’ button

After selecting the desired parameters, the user can click on the “Run the analysis” button. Doing this filters and aggregates if necessary the data, run the signal detection algorithm and updates the graph and table outputs.

5. Time series graph

The time series graph shows the number of cases aggregated by time unit over the study period.

The study period includes:

• The historical data (light green solid line):
  The historical time series on which the signal detection algorithm forecast will be based on (i.e. the estimation of the expected number of cases).
• The signal detection period including:
  • The observed cases (light green dashed line): The actual number of cases reported during the signal detection period (i.e. the forecast period).
    
  • The threshold value computed by the algorithm (dark orange solid line): The upper limit of the expected number of cases (i.e. forecast) estimated with the selected algorithm.
  • The signals detected within the signal detection period (red triangles): Time points where the observed number of cases is higher than the threshold value (i.e. the upper limit of the expected number of cases).

The graph automatically updates when the Run the analysis button is pushed.

6. Result table

The table below the time series shows the date-time of the signal, the geographical unit, the observed number of cases, the corresponding threshold value to the date-time and whether there was an signal.

When EU-EEA complete series is selected, an additional column appears listing all countries included in the EU-EEA aggregated figures and highlighting the observed number of cases by country. This detailed information can help the user to identify which country contribute the most in raising the signal at EU-EEA level.

The table can be sorted by clicking on the column header. A search field is also available at the bottom of each column. The user can use the pagination controls to page through the results of the analysis.

7. Download buttons

The user has several options to download:

• The output result table in csv format;
• The time series plot in PNG format;
• The dataset used to plot the time series in csv format.

8. ‘Generate and download the standard report of signals at country level’ button

Last but not least, the user can generate and download a complete report describing all detected signals at country level. This HTML report summarise the findings of the Signal Detection tool for the given parameters.

Please note that generation of the report can take several minutes.

1. Data quality

Please, always check the quality of your data before jumping into the time series analysis. This tool does not provide any quality checks and relies on a clean dataset.

2. Dataset

Use the default dataset available in the package EpiSignalDetection::SignalData and explore it briefly before jumping into the analysis.

my_dataset <- EpiSignalDetection::SignalData
knitr::kable(head(my_dataset))

3. Setting parameters for the analysis

Set the desired parameters on which you wish to run the analysis on. Please make sure that the selected parameters are available in the given dataset.

my_input <- list(
  disease = "Salmonellosis",
  country = "EU-EEA - complete series",
  indicator = "Reported cases",
  stratification = "Confirmed cases",
  unit = "Month",
  daterange = c("2012-01-01", "2016-12-31"),
  algo = "FarringtonFlexible",
  testingperiod = 5
)

4. Plotting the times series

Once the parameters are defined, you can directly run the plot function as described below.

If you use the default dataset EpiSignalDetection::SignalData, the only required argument is my_input.

EpiSignalDetection::plotSD(input = my_input)

5. Generating the standard report of signals

In order to summarise your findings into an automated report, you can use the following command to generate the standard report of signals.

This command will store it on your computer in a specific folder, for instance C:/R/test.html.

EpiSignalDetection::runEpiSDReport(
  outputfile = "C:/R/test.html", 
  input = my_input)

6. Generating the stratified report of signals

If you wish to stratify your finds by population characteristics (see Population variable), you can use the following syntax.

EpiSignalDetection::runEpiSDReport(
  outputfile = "C:/R/test.html", 
  input = my_input, 
  stratified = TRUE)

Please note that you can also run the report as a standalone function runEpiSDReport(). Each parameter will be then defined through the R console:

EpiSignalDetection::runEpiSDReport()
[1] "Dataset (please enter the full path to the csv file, or just press 'Enter' to use the default dataset)"
1:
Read 0 items
[1] "Disease / Health topic (e.g. Salmonellosis):"
1:  Salmonellosis
Read 1 item
[1] "Region name (e.g. EU-EEA - complete series):"

1. Importing the dataset

Import your own external dataset:

my_dataset <- EpiSignalDetection::importAtlasExport("./data/ECDC_surveillance_data_Pertussis_20180717.csv")

The imported dataset will have to go through the data preparation function cleanAtlasExport.

my_dataset <- EpiSignalDetection::cleanAtlasExport(my_dataset)

2. Setting parameters for the analysis

Set the analysis parameters.

my_input <- list(
  disease = "Pertussis",
  country = "EU-EEA - complete series",
  indicator = "Reported cases",
  stratification = "All cases",
  unit = "Month",
  daterange = c("2012-01-01", "2016-12-31"),
  algo = "FarringtonFlexible",
  testingperiod = 6
)

3. Data preparation

The clean dataset needs to be filtered on the selected parameters.

my_dataset <- EpiSignalDetection::filterAtlasExport(my_dataset, my_input)

4. Data aggregation

Once filtered, the dataset needs to be aggregated by time unit.

my_dataset <- EpiSignalDetection::aggAtlasExport(my_dataset, my_input)

5. Plotting the time series

Finally, you can plot the times series using the syntax below.

EpiSignalDetection::plotSD(x = my_dataset, input = my_input)

6. Generating the standard report of alarms

Last but not least, you can generate the standard report of signals at country level using the command below.

my_input <- list(
  file = list(datapath = "C:/data/ECDC_surveillance_data_Pertussis_20180717.csv"),
  disease = "Pertussis",
  country = "Greece",
  indicator = "Reported cases",
  stratification = "All cases",
  unit = "Month",
  daterange = c("2011-12-01", "2016-12-01"),
  algo = "FarringtonFlexible",
  testingperiod = 3
)

EpiSignalDetection::runEpiSDReport(input = my_input)

7. Generating the stratified report of signals

If you wish to stratify this report by population characteristics, you can use the option stratified = TRUE.

Please note that this report is not available in the Shiny application because its generation is too time consuming for an interactive interface. However, it can be run outside the tool.

my_input <- list(
  file = list(datapath = "C:/data/ECDC_surveillance_data_Salmonella_20180717.csv"),
  disease = "Salmonellosis",
  country = "EU-EEA - complete series",
  indicator = "Reported cases",
  stratification = "Confirmed cases",
  unit = "Month",
  daterange = c("2011-12-01", "2016-12-01"),
  algo = "FarringtonFlexible",
  testingperiod = 6
)

EpiSignalDetection::runEpiSDReport(input = my_input, stratified = TRUE)