CDC’s Division of the Strategic National Stockpile (DSNS) will conduct a webinar on RealOpt at 2:00 PM ET this Wednesday, September 2, 2009. RealOpt is a simulation-based optimization software that can help planners design PODs that run more efficiently and achieve desired throughput rates while operating with fewer staff. Contact Allison Cain at firstname.lastname@example.org for details on how to participate.
We recently worked with a local school district that is planning Flumist clinics in which a batch of 15 students arrives every 5 minutes. You can find the complete clinic planning model that we created online at the Clinic Planning Model Generator software page. The clinic includes a forms check station and a vaccination (flu mist) station.
The President’s Council of Advisors on Science and Technology (PCAST) has released a report assessing the government’s preparations for the resurgence of the H1N1 influenza this fall.
The report describes a “plausible scenario” in which 30 to 50% of Americans are infected, as many as 1.8 million people are admitted to the hospital, 300,000 patients require fill the available intensive care units, and between 30,000 and 90,000 Americans die from the flu.
An interesting point concerns the timing of vaccinations: the report states that the surge in the flu could begin in September and peak in mid-October, but the vaccines may not be available until mid-October.
The report also recommends changes to the decision-making processes so that a single individual is “responsible for coordinating all policy development for the 2009-H1N1 response; identifying the people, agencies, and processes for making key decisions; guaranteeing that all necessary decisions are made in a timely manner; and presenting recommended courses of action to the President.”
The report is online here.
The article by Brandeau et al. in the same issue presents recommendations for mathematical and simulation models:
1. Include stakeholder input to effectively address the spectrum of relevant real-world planning and response problems.
2. Include a user-friendly interface, the ability to customize model inputs to suit local needs, the ability to quickly and easily perform sensitivity analysis, and provide ongoing user support.
3. Build models that balance simplicity and complexity so that our models both adequately represent real-world scenarios and can be used and interpreted by the intended end users.
4. Include in our models the relevant outcomes, including outcomes that extend beyond those typically considered in cost-effectiveness analyses; these may include timeliness and efficiency of response, resource utilization, evacuation timing, and/or measures related to behavior of responders and the public.
5. Build models that address the fundamental uncertainties in disaster scenarios, including the likelihood and magnitude of an event, operational response capabilities, supply chain capacity and robustness, behavior of responders and the public, and countermeasure effectiveness.
6. In our reports, address the motivation for the study, define critical assumptions, explain the modeling methodology, discuss key sensitivity analyses, make available public-use versions of the model, and identify
all relevant partners in model creation.
The July/August 2009 issue of Medical Decision Making includes an article by Nathaniel Hupert and his colleagues describing a model that determines the expected number of individuals who develop symptomatic inhalational anthrax and require hospital-based intravenous antibiotic treatment. The key parameters in their study are delays in the dispensing of antibiotics and the effectiveness of the antibiotics. They conclude that extending the duration of the dispensing has less impact than delays in the start of dispensing or reductions in antibiotic effectiveness, both of which can dramatically increase the number of people who need to go to the hospital.
The model is a discrete-time model implemented in a Microsoft Excel workbook with a VBA macro.
The Montgomery County, MD, Advanced Practice Center for Public Health Emergency Preparedness and Response and I would like to remind everyone that POD planning tools like the Clinic Planning Model Generator (CPMG) can be used to plan staffing and predict queueing and throughput H1N1 vaccination in PODs and in schools.
Some of these will be very simple – for instance, a school POD (clinic) may have just a vaccination station staffed by nurses. The inputs to the CPMG are the number of students at a school and the number of hours for the vaccination activity. The model will determine how many staff are needed. If the number needed exceeds the number available at the school, then increase the number of hours that the POD will operate.
If you have any questions, let me know at email@example.com.
(August 4, 2009) UPDATE: See also the CDC’s list of Tools and Models to Estimate Staffing.