PUBH 8546 Week 8 Threshold Memo Example

Course	PUBH 8546 Advanced Analysis of Community Health Data and Surveillance in Public Health
Week	Week 8
Paper type	Threshold memo
Length	About 660 words, 4 pages
Format	Memo, APA 7 citations
School	Walden University
Program	PhD in Public Health
Updated	October 2026

Free sample paper for PUBH 8546 Week 8

Memorandum

To: Surveillance Epidemiologist, Northwater County Health Department

From: Doctoral Student, Community Health Data and Surveillance

Date: Month Day, Year

Re: Alert thresholds for weekly cold-related emergency visits this winter

Purpose

This memo sets the rules by which a weekly count of suspected cold-related visits across Northwater's emergency departments will be judged unusual this winter. Two kinds of rise matter. A sudden jump, such as a cluster during a cold snap or among people sleeping outside after a shelter closes, needs a limit for single weeks. A gradual climb over several weeks needs a method that accumulates small excesses. The rules below are fixed before the season starts.

What this page is doingNaming both kinds of rise first explains why the memo sets two rules rather than one.

The Baseline

The baseline for each week is the mean count for the same calendar week and the two weeks on either side of it in the previous three winters, giving fifteen weeks of history. Using the same time of year matters because cold illness rises and falls with the season; a January week compared with November weeks would trigger alerts every year. Only facilities that reported throughout the baseline period are included, and weeks in which a facility joined or left the feed are excluded. A guard band leaves out the two most recent weeks of the current season, so that a rise already under way does not raise its own baseline and hide itself, a design choice used in widely deployed aberration detection methods (Hutwagner et al., 2003).

The Sudden Jump Rule

With an illustrative baseline mean of 10 visits in a midwinter week, a Poisson distribution gives the one-sided 99% limit as 18. A week with more than 18 visits would occur less than 1% of the time if nothing had changed. Such a week triggers same-day review.

Table 1

Alert rules for the coming winter

Rule	What it detects	Setting	Action when crossed \|
Poisson limit	A single unusual week	Above the 99th percentile of the weekly baseline	Same-day review of records and facilities \|
Cumulative sum	Several modestly high weeks in a row	Reference value half a standard deviation above baseline; decision limit four standard deviations	Review within two days; check for gradual artifacts \|
Facility share check	One hospital driving an alert	Any facility contributing over half of the week's excess	Contact that facility before any public message \|

Settings are illustrative design choices for a composite county.

The Gradual Climb Rule

A count of 13 or 14 visits a week would never cross the Poisson limit, yet three such weeks in a row could signal a real rise. A cumulative sum adds each week's excess above a reference value and signals when the running total passes a decision limit; it resets toward zero when weeks fall back below the reference. Comparisons of detection methods have found cumulative sum approaches better at detecting gradual increases than single-week limits (Fricker et al., 2008). Here the reference value is set half a standard deviation above the baseline mean and the decision limit at four standard deviations, a common starting point that the department can tune after the season.

The Figure

A single figure accompanies each weekly report: the weekly count plotted over the winter with the baseline mean and the 99% limit drawn as lines, and the cumulative sum plotted beneath it with its decision limit.

Why Fix the Rules Now

A threshold chosen after seeing a high week is no longer a threshold; it becomes a judgment shaped by the result. Syndromic systems exist to detect change early (Henning, 2004), and that purpose is served only if the definition of unusual is settled in advance. Should a rule change mid-season, the change, its date and the reason go into the log. After the season, the department will review how many alerts each rule produced and how many proved real.

Decision Requested

Please approve the baseline, both rules and the facility share check before the first week of December.

References

Fricker, R. D., Hegler, B. L., & Dunfee, D. A. (2008). Comparing syndromic surveillance detection methods: EARS' versus a CUSUM-based methodology. Statistics in Medicine, 27(17), 3407-3429. https://doi.org/10.1002/sim.3197

Henning, K. J. (2004). What is syndromic surveillance? (MMWR Supplements, Vol. 53). Centers for Disease Control and Prevention.

Hutwagner, L., Thompson, W., Seeman, G. M., & Treadwell, T. (2003). The bioterrorism preparedness and response Early Aberration Reporting System (EARS). Journal of Urban Health, 80(Suppl. 1), i89-i96. https://doi.org/10.1007/PL00022319

Reading the PUBH 8546 Week 8 instructions

Threshold or aberration detection assignments ask students to define how their surveillance system will decide that a count is unusual. Prompts usually require a baseline method, one or more detection rules, the settings for each, and the actions that follow an alert. Many instructors ask students to address seasonality and to explain why rules are fixed before data arrive. Memos that rely on a single fixed number without a baseline, or that compare winter weeks with summer weeks, generally lose credit, because such rules produce predictable false alarms or miss the gradual rises that matter most.

How this PUBH 8546 Week 8 example is built

The memo opens by naming two kinds of rise, sudden and gradual, and explaining why each needs its own rule. A baseline section describes matching calendar weeks from three earlier winters, facility restrictions and a two-week guard band. A section derives a 99% Poisson limit from an illustrative baseline mean. A rules table lists three rules with settings and actions. A section explains the cumulative sum and its parameters, citing a comparison of detection methods. A short section describes the weekly figure, and a closing section explains why the rules are fixed now and how changes will be recorded. Settings in the rules table are labeled as illustrative design choices, and the memo promises a post-season review of how many alerts each rule produced and how many proved real. A request for approval before December closes the memo.

Reading the PUBH 8546 Week 8 rubric

Graders typically evaluate the soundness of the baseline, appropriateness of the detection rules, clarity of settings and actions, and commitment to pre-specification. A seasonal baseline from matching weeks earns credit. A guard band shows awareness that rising counts can mask themselves. Pairing a single-week limit with a cumulative method demonstrates understanding of different signal shapes. A facility share check shows practical sense about artifacts. Memos that state actions for each rule and commit to post-season review usually score better than those that stop at statistical settings. Clear numerical settings, rather than descriptions such as a high count, make the rules testable and repeatable. Graders also reward memos that explain the cost of false alarms and missed signals, since thresholds always trade one against the other.

PUBH 8546 Week 8 help: mistakes that cost marks

Decide which past weeks represent normal for each week you will monitor, accounting for season. Leave out the most recent weeks so a rising trend does not hide itself. Choose a rule for sudden jumps and a rule for gradual rises, and write their settings as numbers. Add a check for a single facility driving an alert. Write down what happens when each rule is crossed. Fix everything before the season and log any change. If you are unsure which settings suit your counts, we can draft a sample threshold memo for your data.

Get a PUBH 8546 Week 8 example written to your instructions

Include the Week 8 memo instructions, the rubric and a sense of how often your indicator is counted. The memo fixes a baseline, a jump rule, a slow-climb rule and a response path before any data are read, and comes back in 24-48h, a first sample costing nothing. Its weekly series is simulated; the thresholds your indicator needs depend on its own history.

More PUBH 8546 papers and PhD in Public Health samples

PUBH 8546 Week 8 questions, answered

Where can I find a free PUBH 8546 Week 8 sample paper?

The full Week 8 threshold memo is shown above, setting alert rules for cold-related visits.

How is the baseline built?

From the same calendar week and the two weeks on either side in three previous winters.

What is the guard band for?

Leaving out the two most recent weeks so a rise already under way does not inflate its own baseline.

What is the Poisson limit for a baseline mean of 10?

Eighteen; a week above that would occur less than 1% of the time if nothing had changed.

Why add a cumulative sum?

It accumulates modest excesses over several weeks and detects gradual climbs that single-week limits miss.