Every Number in This Blog Has a Link. That Was a Deliberate Choice.

Every Number in This Blog Has a Link. That Was a Deliberate Choice.

Here is how a statistic dies in home health marketing. A vendor publishes a post claiming that some share of referrals never convert. A second vendor cites the first. A trade publication cites the second. A conference deck cites the trade publication. Eighteen months later the figure is everywhere, attributed to nobody, and no longer traceable to a dataset, a methodology, or a year.

Nobody lied. The number just stopped having a source, and once that happens you cannot tell a real finding from a plausible one.

We decided not to participate in that. This post is the editorial policy behind this library, written down so you can hold us to it.

The house rules

These are the healthcare content sourcing standards every post here is written against.

• Every figure is an inline link to the source it came from. Not a bibliography at the bottom, not "according to CMS" — a hyperlink on the number itself, pointing at the page where you can see it. If you cannot click through to the figure, we should not have printed it.
• The source has to be primary. Statute and regulation text, federal rulemaking documents, agency datasets, agency reports, government audit reports, and named peer-reviewed journals. Not a vendor blog, not a webinar recap, not another article citing a primary source. If the only path to a number runs through someone else's content, the number does not go in.
• The source has to have been opened. A citation is a claim that someone read the page. If the page cannot be loaded, the figure is dropped, even when it is almost certainly correct and widely repeated.
• Missing numbers are acceptable. Unverifiable numbers are not. A point made qualitatively is weaker than a point made with a verified figure and infinitely stronger than a point made with a figure nobody can check.
• No regulation gets rounded up. Plenty of vendor content claims a rule requires call recording. Very few rules require call recording. We describe what a rule literally says and what it literally covers, which is often narrower and less convenient than the marketing version.
• Contested law gets presented as contested. Where a question genuinely varies by state or is unsettled, the post says so and cites both directions rather than picking the reading that suits us.
• No proof claims. Nothing here cites our own customers, outcomes, or results. Illustrative scenarios are labeled as hypothetical. A vendor's internal data is not evidence you can check, so it does not belong in an argument we are asking you to evaluate.

A worked example, including the part that was inconvenient

Policies are cheap. Here is one that cost something.

A brief for a recent post in this library carried a figure from a 2024 study in The American Journal of Managed Care on home health referrals and readmissions: roughly 24% versus 16% at 90 days. The framing in the brief was the intuitive one, the one that writes itself for a company selling call intelligence to home health agencies — unfulfilled referrals lead to more readmissions, so agencies should answer the phone.

Loading the study showed the groups were the other way around. The patients who received home health had the higher 90-day readmission rate, 24% against 16%, and the lower mortality, 8% against 10%. The convenient version of that sentence was not what the authors found.

Three things followed from that, and they are the actual argument for this policy:

• The figure got published in the direction the study reported, not the direction that helped.
• The surrounding claim had to change. The post no longer argues that answering referral calls reduces readmissions, because that study does not support it.
• The caveats went in. Observational design, a convenience sample from a single state, Medicare Advantage only, unmeasured confounding possible. The authors say all of this plainly, and a reader deciding what to do on Monday deserves to know it.

If we had taken the number from a secondary summary, we would have inherited the reversal and never known. That is the whole case for opening the source.

What the policy costs

It is worth being honest that this makes for worse marketing content in several measurable ways.

What the rule prevents	What it costs
No figures we could not load	Posts ship with fewer statistics, and sometimes with qualitative statements where a number would land harder
Primary sources only	Research takes considerably longer, and some good arguments get abandoned because the underlying data is paywalled or does not exist
No rounding up a regulation	We cannot write the urgency-driven compliance content that performs best, because most rules are narrower than that content implies
Contested law stays contested	Posts end in judgment calls for the reader rather than clean directives
No customer proof	No case studies, no outcome metrics, no ROI figures — the three things buyers ask for most

We think that trade is correct for this audience. Compliance officers and operations leaders in home health, home care, and hospice are the people who get asked to defend a number in front of a surveyor, a board, or an auditor. A figure you cannot source is a liability to them, not an asset.

How to audit any vendor's content in ten minutes

Apply this to us, and to everyone else sending you content this quarter.

1. Click three numbers

Pick the three most impressive statistics in a post. Click them. You are checking whether the link exists, whether it resolves to a primary source rather than another article, and whether the number is actually on the page it points to. Dead links and circular citations show up immediately.

2. Read what the regulation says, not what the post says it says

When a post claims a rule requires something, open the cited section and read the operative sentence. The frequent failure is scope: a requirement that applies to one program, one plan type, or one category of marketing gets described as though it applies to everyone. Our own posts on what retention rules actually apply to call recordings and on why the common all-party consent state list is wrong exist because both claims circulate in overstated form.

3. Check the year and the denominator

A current-sounding figure is often four years old, and a percentage without a denominator is not a finding. If the post does not tell you the population and the period, the source page should. If neither does, discard it.

4. Look for the caveats

Observational studies, convenience samples, and single-region cohorts are not disqualifying — most useful healthcare research has at least one of these. But a post that cites such a study without mentioning its limits is telling you something about its own standards.

5. Ask what the vendor cannot say

Content that has no acknowledged limits, no contested questions, and no dropped claims has not been edited. It has been produced.

Why this matters more for operational content than most

The reason to be strict here is that this library is largely about what you should do differently on Monday. Content that changes operational behavior in a regulated setting carries a different obligation than content that fills a newsletter. If a post sends an agency to restructure its intake coverage or expand its call review scope, the evidence behind that recommendation should be something the reader can inspect and disagree with.

That is the same reason we keep arguing for surfaces over dashboards in operational reporting: a number you cannot trace to the thing that produced it is a number you cannot act on with any confidence, whether it appears in a blog post or on an executive report.

SurfacerIQ builds in the category of conversation intelligence and compliance monitoring for healthcare contact centers, which means our product is also in the business of making a figure traceable back to the specific interactions underneath it. The editorial standard and the product standard are the same standard, and it would be strange to hold one and not the other.

If you think a figure in this library is wrong, misframed, or missing its caveats, tell us and we will correct the post. Talk to us.

This post and the regulatory content it describes are general information for operators, not legal advice. Specific obligations turn on your program, state, payer mix, and facts, and should be reviewed with your own counsel or compliance officer.