Nursing home antibiogram software for long-term care antibiotic stewardship

Four ways long-term care facilities reach a usable isolate count

The published approaches to this problem are not exotic. A 2018 review in the Journal of the American Medical Directors Association set out the options a facility with few annual isolates actually has, and every one of them is a decision about scope rather than a new data source.

The first, pooling more than one year of cultures, is the one most facilities should take, and it is the one the Microbiology Spectrum simulation supports directly: aggregating two years measurably reduced mean error at every sample size tested. The cost is that a two-year antibiogram smooths a trend, so it tells you what to start empirically but not whether resistance moved last quarter.

A nursing home attached to a critical access hospital has a fifth option: pool with the hospital's swing bed and inpatient cultures and keep each site as its own column. That setup is described on the critical access hospital antibiogram page. The proxy option works poorly when your residents come back from a long-term acute care hospital, whose resistance is far higher than the short-stay hospital's; that setting has its own LTACH antibiogram page.

• Extend the analysis window beyond twelve months, usually to two or three years, with the window stated on the report
• Combine isolates across facilities in the same region or the same operator group, so a chain of six homes reports as a network
• Use a nearby acute-care hospital antibiogram as a proxy, accepting that its population is sicker and more heavily exposed
• Collapse the report: fewer organism rows, fewer drug columns, grouped by the decision being made rather than by the full panel

Where the antibiogram sits in CMS F881 and the CDC Core Elements

The federal requirement that puts this on a director of nursing's desk is 42 CFR 483.80(a)(3), cited on survey as F881. It requires an antibiotic stewardship program inside the facility's infection prevention and control program, with protocols and a system of monitoring antibiotic use. Surveyors investigating a stewardship concern work through the unnecessary medication review pathway for at least one resident on an antibiotic, so what a facility can put in front of them matters.

The CDC Core Elements of Antibiotic Stewardship for Nursing Homes name seven elements: leadership commitment, accountability, drug expertise, action, tracking, reporting and education. Two of those seven are where an antibiogram is the cheapest available evidence. Tracking asks the facility to monitor at least one process measure and one outcome of antibiotic use. Reporting asks for regular feedback on antibiotic use and resistance to prescribing clinicians and nursing staff. A dated, facility-specific susceptibility report with its rules printed on it satisfies the resistance half of both, and it does so with data the reference lab is already producing.

AHRQ went further and published a nursing home antibiogram program toolkit inside its antimicrobial stewardship guide, structured as four phases: assessment and planning, development, implementation, and program monitoring. The toolkit is free and worth reading. What it does not do is build the report, which is the part that has historically cost a consultant pharmacist a fortnight in a spreadsheet.

The infection tracking side of the same program, and which products build what, is compared in antibiotic stewardship software for nursing homes.

Start with urine, because that is most of what you have

Urinary tract infection is the most common reason a resident is started on an antibiotic in long-term care, and urine is the specimen a nursing home actually sends in volume. That makes urine the one source where a facility has a realistic chance of clearing 30 first-isolate Escherichia coli results within a two-year window, and it makes the urine antibiogram the one that changes prescribing.

It also concentrates the drug choice. Nitrofurantoin, trimethoprim-sulfamethoxazole, cephalexin and a fluoroquinolone are the agents a nursing home guideline argues about, and the argument is settled by local percentages, not by a textbook. Nitrofurantoin is worth reporting per organism rather than pooled, because it is a reasonable choice against Escherichia coli and a poor one against Proteus mirabilis, and a single pooled nitrofurantoin figure hides that entirely.

One caution that belongs on every long-term care antibiogram: asymptomatic bacteriuria is common in this population, so a proportion of the isolates behind the report came from residents who were colonized rather than infected. That does not invalidate the susceptibility percentages, which describe the organisms, but it is a reason the report should never be read as a licence to treat a positive culture on its own.

What to ask your reference lab to send

Nursing homes rarely have an on-site microbiology laboratory, so the export comes from a reference lab such as a regional hospital lab or a national provider. The request is small and every lab can fill it. Ask for a line-level CSV of finalized culture and susceptibility results for your facility, covering the last 24 to 36 months, with one row per organism and drug result.

Seven fields carry the whole antibiogram: a resident identifier that can be pseudonymous but must be stable, the collection date, the specimen source, the organism name, the antibiotic, the interpretation of S, I or R, and the MIC where one exists. If the lab can add a unit or wing code, take it. If it cannot, the facility-level report is still the report you are required to produce.

Ask for the breakpoint version too, in writing. A susceptibility percentage that appears to improve between two reporting periods sometimes reflects a revised CLSI M100 breakpoint rather than any change in the organisms, and a facility that cannot say which table was applied cannot explain its own trend to a surveyor or to a prescriber.

Reference labs that are asked for this by many facilities at once can see how the lab side works in antibiogram software for reference labs.

• Resident identifier, stable across the window, so first-isolate de-duplication works
• Collection date, to set the pooling window and to sort repeat isolates
• Specimen source, so the urine cut can be made
• Organism, antibiotic, and the S, I or R interpretation
• MIC value where the method produced one
• The breakpoint table version applied, stated once per export

What the finished report should say on its face

A long-term care antibiogram is read by people who were not in the room when it was built, sometimes a year later, sometimes a surveyor. Everything needed to interpret it has to be printed on it. That means the isolate count under every percentage, the pooling window in plain words, the de-duplication rule, how intermediate results were handled, the breakpoint set and its version, and a visible mark on every cell that fell below the reporting threshold.

Suppressed cells are not a failure. A report that says n equals 12 and declines to give a percentage is more useful, and considerably more defensible, than one that prints 58 percent from twelve isolates and invites a prescriber to act on it. Antimicrobe suppresses below the threshold you set and shows the count anyway, so the reader can see the gap and the facility can see which organism is one year of pooling away from being reportable.

The practical test is whether a consultant pharmacist who has never seen the file can pick up the report and explain every number on it. If they can, the report will survive a survey, a P and T discussion, and next year's comparison.