How Hospitals Use Clinical Benchmarking to Set Realistic Quality Improvement Targets: An American Data Network Guide

Consider a Quality Director who reviews the latest clinical benchmarking report and sees the hospital’s 30-day heart failure readmission rate sitting well above its peer group median. Leadership has seen the same gap, so the Director sets the goal that seems most defensible: reach the peer median within 18 months. The care transitions team adds post-discharge follow-up calls and tightens medication reconciliation, and the rate improves.

Eighteen months later, the hospital has still missed its target. The initiative did not fail. The target never accounted for the hospital’s starting point, its case mix, or how fast the measure could realistically move.

A clinical benchmarking gap and an improvement target are not the same thing. The gap shows where a hospital stands relative to peers. It does not show where the hospital should aim, how fast it can move, or what level of improvement is worth the resources required. When the two are treated as one, hospitals may launch initiatives against goals they cannot reach with the levers available, which can lead to initiative fatigue or misallocated resources.

Closing that distance takes a structured step between gap and goal: validate the gap, then size the target against what the hospital has actually shown it can move. This guide from American Data Network (ADN) walks through both steps.

Key Takeaways

• A clinical benchmarking gap is a starting point for target setting, not the target itself. Peer medians and round numbers make easy targets but often ignore a hospital’s starting point and capacity to change.
• Before a gap becomes a target, validate it: confirm the peer group is comparable, the gap is statistically reliable given case volume, and the difference reflects care rather than documentation.
• Peer group assignments shift. In CMS’s 2026 Overall Star Ratings, the newly added OAS CAHPS measures placed 98 hospitals in Peer Group 4 that otherwise would have been in Peer Group 3.
• A defensible quality improvement plan target accounts for three inputs: the hospital’s demonstrated rate of change, the operational levers actually available, and the resource cost relative to expected benefit.
• Governance reporting should give leaders the baseline, peer group limits, target rationale, and progress they need to make decisions without clinical detail.

What Does a Clinical Benchmarking Gap Actually Tell You?

A gap on a clinical benchmarking report is an observation, not a conclusion. Three checks help determine whether it is solid enough to build a target around.

Confirm the Peer Group Is Comparable

A gap only means something if the comparator hospitals resemble yours in size, service mix, and case mix. Peer group assignments can also shift from one cycle to the next.

Under the Overall Star Rating methodology from the Centers for Medicare & Medicaid Services (CMS), hospitals are compared within peer groups based on how many measure groups they report. In CMS’s analysis of the 2026 ratings, the newly added Outpatient and Ambulatory Surgery Consumer Assessment of Healthcare Providers and Systems (OAS CAHPS) survey measures placed 98 hospitals in Peer Group 4 that otherwise would have been in Peer Group 3.

ADN covers peer group selection in depth in its guide to avoiding misleading benchmarking comparisons, and the severity-adjusted data in its Clinical Benchmarking Application supports comparisons at the facility, service line, and physician levels.

Check Reliability Against Case Volume

Small denominators can increase uncertainty around healthcare quality metrics, and CMS applies measure-specific reporting thresholds, such as minimum case counts, before publicly reporting a hospital’s score. CMS reports measures such as 30-day pneumonia mortality as risk-standardized rates with interval estimates. At a hospital with few eligible cases, greater uncertainty around the estimate may warrant additional review before treating an apparent difference from peers as a confirmed performance gap.

Rule Out Documentation and Measurement Effects

Documentation and coding practices can shape how performance appears in the data, and what to review depends on how the measure is built. For chart-abstracted measures, a review can check whether abstraction, including exclusions, followed the measure specifications.

Abstraction accuracy is part of that check. ADN’s clinical data abstraction services maintain 98.4% accuracy across all measures and registries, all abstractors, and all clients, supported by a rigorous Inter-Rater Reliability (IRR) program.

For claims-based measures such as CMS’s 30-day heart failure readmission measure, coded diagnoses and procedures carry more weight: they help determine which stays qualify as index admissions, and CMS’s planned readmission algorithm uses them to classify readmissions as planned or unplanned.

Reviewing a sample of the cases behind a measure against its published specifications can help determine whether a gap reflects care delivery rather than how care was recorded.

Only a validated gap should become a target. A gap that fails one of these checks may still deserve attention, but the first step is often fixing the comparison or the documentation rather than launching an improvement initiative.

How Do You Set a Target That Is Both Ambitious and Achievable?

Once a gap is validated, a defensible target in a quality improvement plan rests on three inputs, followed by one common trap to avoid.

Anchor the Target to Demonstrated Rate of Change

Start with how fast this measure has actually moved at this hospital before. How much has it changed over the past several reporting periods, and in which direction? What improvement did the hospital achieve the last time it ran a comparable intervention over a similar time frame? Does the proposed target require a rate of change well beyond anything the organization has demonstrated? If it does, the target likely needs more time, more resources, or a nearer-term interim goal.

Answering those questions depends on a clean historical view of the hospital’s own healthcare quality metrics. ADN’s healthcare data analytics services analyze hospital data to identify patterns, trends, and priorities across clinical, quality, financial, and patient safety areas.

Map the Operational Levers Available

A target is only as realistic as the changes behind it. For a heart failure readmission gap, the levers might include discharge education, follow-up call capacity, medication reconciliation, and post-acute coordination. Leaders should confirm which of those the Quality team controls and which require leadership approval or budget. Testing a lever on a small scale through Plan-Do-Study-Act (PDSA) cycles before building the full target around it can show how much movement it realistically produces.

Weigh Resource Cost Against Expected Benefit

A useful estimate covers staff time, including any dedicated full-time equivalents (FTEs); direct intervention costs, such as follow-up call staffing or patient education materials; and the time horizon before results are likely to show in the data. On the benefit side, it covers the expected quality gain and any financial or public reporting exposure tied to the measure. Setting the two side by side helps leaders decide whether a given level of improvement is worth pursuing now or should be phased.

Avoid Default Targets

Peer medians and round numbers, such as “reach 90%,” are appealing because they are easy to explain. But neither reflects where the hospital starts or how fast it can move. The peer median can remain a reasonable long-term reference point while the near-term target in the quality improvement plan is set from the hospital’s own demonstrated rate of change.

Benchmarks also move over time, so a peer median chosen today may shift by the target date, a dynamic ADN explores in its look at national patient safety benchmarks and shifting baselines.

What Happens When Target Setting Skips This Analysis?

When a target skips validation, a review of the improvement pathway, or resource planning, a successful initiative can still register as a failure, just as the heart failure initiative did.

That distinction matters when leaders decide what happens next. A missed target can suggest that an intervention should be changed, scaled back, or discontinued even when the underlying measure moved meaningfully in the right direction. It can also make it harder to determine whether the problem was the improvement strategy or an unrealistic expectation of how much change was achievable within the time and resources available.

Repeated across initiatives, that pattern can produce initiative fatigue among frontline teams, frustration among leaders who funded the work, and declining confidence in the quality improvement process itself, even when real improvement occurred.

How Do You Make a Quality Improvement Plan Target Governable?

A target built from clinical benchmarking data also has to hold up in governance review. ADN’s guide to what boards need from hospital quality reporting covers the broader communication framework. For a specific target, leaders typically need to see:

• Baseline data across several reporting periods, with the measurement period and data source noted, so leaders can see whether the starting point is stable.
• The comparator peer group, how it was selected, and its known limits, such as differences in size or case mix or a recent change in group assignment.
• The rationale for the target level, tied to the hospital’s demonstrated rate of change and available resources, along with interim milestones.
• A consistent follow-up view at each quality committee or board review, showing performance against the current milestone, trend since baseline, intervention status, and any decision leadership needs to make.

Keeping the view focused on those decisions, rather than on clinical detail, helps leaders govern the target instead of simply receiving updates on it.

Building Targets That Hold Up

Healthcare quality metrics are most useful when they inform a plan a hospital can actually execute. A target built on a validated gap and the hospital’s own capacity to change is one Quality leaders can defend and teams can reach. Clinical benchmarking is where that work starts, not where it ends.

ADN’s Clinical Benchmarking Application supports that starting point with monthly-updated clinical, quality, and financial information, peer and national benchmarks, and a Quality Module covering more than 160 quality and patient safety indicators, including those endorsed by the Agency for Healthcare Research and Quality (AHRQ) and the National Quality Forum.

ADN’s Data Analytics Services supply the other half: the trend analysis behind a target’s demonstrated rate of change. For a broader view of how analytics supports quality work, see ADN’s guide to the benefits of data analytics in healthcare.