Dynamic Interactive Educational Diabetes Simulations Using the World Wide Web: An Experience of More Than 15 Years with AIDA Online

Abstract

Background. AIDA is a widely available downloadable educational simulator of glucose-insulin interaction in diabetes. Methods. A web-based version of AIDA was developed that utilises a server-based architecture with HTML FORM commands to submit numerical data from a web-browser client to a remote web server. AIDA online, located on a remote server, passes the received data through Perl scripts which interactively produce 24 hr insulin and glucose simulations. Results. AIDA online allows users to modify the insulin regimen and diet of 40 different prestored "virtual diabetic patients" on the internet or create new "patients" with user-generated regimens. Multiple simulations can be run, with graphical results viewed via a standard web-browser window. To date, over 637,500 diabetes simulations have been run at AIDA online, from all over the world. Conclusions. AIDA online's functionality is similar to the downloadable AIDA program, but the mode of implementation and usage is different. An advantage to utilising a server-based application is the flexibility that can be offered. New modules can be added quickly to the online simulator. This has facilitated the development of refinements to AIDA online, which have instantaneously become available around the world, with no further local downloads or installations being required.

Figure 1

Basic structure of AIDA online. HTML = HyperText Markup Language. The AIDA online homepage (http://www.2aida.net) presents the user with various simulator options and case scenarios. A series of dedicated (Common Gateway Interface) cgi-bin scripts written in Perl v5.0 are used to read case scenario data and insulin and carbohydrate profiles from various databases. The plasma insulin and blood glucose (BG) profiles are computed using a further Perl script which contains the AIDA v4 model differential equations (and which makes use of some temporary storage space on the AIDA online server) [11]. Output from the simulator is returned to the user in HTML format for display by a web browser.

Figure 2

(a) Data from an ASCII (American Standard Code for Information Interchange) text file for an example case scenario (number 0010: Hugh Allibaster) from the AIDA online database. Depending which case scenario the user selects, these data are read in and used as the basis for an initial (baseline) simulation. Forty case scenarios exist in AIDA online, data imported from AIDA v4. Colour key, highlighted in figure. Yellow: weight (in kg). Green: meal times and grams of carbohydrate. Cyan: preparations, times, and dosages of insulin injections. Magenta: renal threshold of glucose, creatinine clearance rate, and hepatic and peripheral insulin sensitivity parameters, respectively. (b) Hidden fields in the dynamically generated HTML web pages produced by the AIDA online cgi-bin Perl scripts. These fields are used to store parameters and case details to be transferred between simulation runs. “units_select” value = 18 for mg/dL blood glucose units; value = 1 for mmol/L. Premixed/biphasic “insulin_type” value = 2. “display” value = 1 is standard display; “display” value = 2 is advanced/fluxes display. “run” value = 1 is the simulation run number (1 = first, baseline simulation). “run” value number increments for each subsequent simulation. “bounds_limits” value = 1 would show the user-defined normoglycaemic ranges. 5 digit number is a PID = process identification number generated by the web server for each separate process/simulation.

Figure 3

(a) AIDA online² main options.htm Case Selector screen shown accessed via Internet Explorer on a Windows XP personal computer. The display shows how a list of summaries of each case scenario can be accessed via a pull-down menu. In this example, a NEW CASE is highlighted. (b) AIDA online² intro.cgi dynamically generated HTML page created based on the NEW CASE selected shown in (a). As can be seen, all the fields are blank, allowing the user to create a new case from scratch. (c) AIDA online² cases.htm fixed HTML web page giving further details of each of the first ten AIDA online case scenarios (imported from AIDA v4).

Figure 4

(a) AIDA online² options.htm Case Selector screen showing Case Scenario 0010 highlighted. Options to select the “Blood Glucose Units” and “Standard” versus “Advanced (includes fluxes)” display are shown lower down the page. (b) AIDA online² dynamically-generated HTML web page, produced by intro.cgi Perl script, populated with information and data for case scenario number 0010 (Hugh Allibaster). Clicking on the (Run Simulation) button at the bottom of the web page will submit the data shown to the simulate.cgi script to run the simulation engine and generate a graphical simulation (as shown in Figure 4(c)). (c) Baseline simulation using AIDA online² with data shown in Figure 4(b) for Hugh Allibaster (case scenario 0010). Top panel: Blood glucose (BG) level and carbohydrate intake over a 24 hour period. User definable normoglycaemic ranges (“Upper/Lower Bounds”: 4–10 mmol/L [72–180 mg/dL]) are shown superimposed. Lower panel: Plasma insulin level and injections of Actrapid (short-acting) and UltraLente (long-acting) insulins. The HbA_1c value (8.6%) gives an indication—for educational purposes—of the glycosylated haemoglobin level predicted by the AIDA model, if the current glycaemic control were to be maintained in the medium- to long-term. (d) Scrolling down the simulate.cgi page shown in Figure 4(c) yields a dynamically-generated HTML data entry form populated with the data used for the simulation, which—for a baseline (initial) simulation—will be identical to that shown in Figure 4(b). The user can now change any of the data values and run a further simulation by clicking on the (Run Simulation) button at the bottom of the screen.

Figure 5

(a) AIDA online² simulation of the case shown in Figure 4(c) in which the evening (7:30 pm) Actrapid dose has been increased by 2 units from 4 units to 6 units. The current simulation run is shown as the vertical red lines, with the previous run shown for comparison as the solid blue lines. The higher plasma insulin peak in the evening and the reduction in blood glucose (BG) overnight are clear to see, although the BG profile still remains elevated compared with the user defined normoglycaemic range. The glycosylated haemoglobin (HbA_1c) level—if this control were maintained in the medium-to-long-term—is expected to improve from 8.6% (Figure 4(c)) to 8.3%. (b) AIDA online² simulation of the case shown in (a) in which the evening (11 pm) Ultralente dose has been increased by 2 units from 8 units to 10 units, resulting in some further improvement in the blood glucose (BG) profile, with an expected glycosylated haemoglobin (HbA_1c) of 7.9% (improved from 8.3% in (a)). The current simulation run is shown as the vertical red lines, with the previous run shown for comparison as the solid blue lines. (c) AIDA online² simulation of the case shown in (b) in which the bedtime snack has been decreased by 5 grams from 10 grams to 5 grams and supper has been decreased by 10 grams from 40 grams to 30 grams (in one simulation). This results in further improvement in the simulated blood glucose profile with a predicted glycosylated haemoglobin (HbA_1c) of 7.6%. Once again, the current simulation run is shown as the vertical red lines, with the previous run shown for comparison as the solid blue lines.

Figure 6

(a) Advanced Display option showing fluxes for the baseline simulation from Figure 4(c). “Glucose Absorption Rate” shows glucose absorption following carbohydrate ingestion in meals; “Renal Excretion” shows the passage of glucose into the urine, above the renal threshold of glucose; “Peripheral Glucose Utilization” demonstrates the utilization of glucose in the periphery, while “Net Hepatic Glucose Balance” shows the production or utilization of glucose by the liver. (b) Screenshot from AIDA online² showing two of the entries in the online glossary—directly accessible at http://www.2aida.org/glossary—which is linked to the online simulator output. In this way, users can click on HTML links to obtain explanations about terms they do not understand or concepts with which they are unfamiliar.

Figure 7

Independent user reviews of AIDA online, reproduced from the AIDA online website (http://www.2aida.org/aida/online-reviews.htm).

Figure 8

Usage of AIDA online showing the number of simulations run per month versus time. (a) January 1999—server hosting “AIDA online” version 1 crashed, and therefore the facility was unavailable for much of the month. (b) October 2000 to January 2001—simulated blood glucose data harvested from AIDA online by researchers at NASA for training and testing an insulin-dosage adjustment decision support prototype. Over 68,000 simulations were run during this time. Further information available at http://www.2aida.org/nasa.

Figure 9

(a) Gives a baseline simulation for case scenario 0026 in the AIDA online database, showing a premixed (biphasic) insulin injection regimen with Mixtard 30/70 being injected twice a day. Mixtard 30/70 is a premixed mixture of 30% short-acting insulin and 70% intermediate-acting insulin. (b) Demonstrates the effect of switching the insulin type from Mixtard 30/70 shown in (a) to Humulin 50/50 (50% short-acting and 50% intermediate-acting insulin) with proportionally more short-acting insulin. The current simulation run is shown as the vertical red lines, with the previous run shown for comparison as the solid blue lines. (c) Demonstrates the effect of switching the insulin type from Humulin 50/50 shown in (b) to PenMix 10/90 (10% short-acting and 90% intermediate-acting insulin) with proportionally more intermediate-acting insulin. The current simulation run is shown as the vertical red lines, with the previous run shown for comparison as the solid blue lines.