Reporting principles

Every figure names its cohort. A rate without a denominator is not an outcome; it is a claim.
Figures from different studies are labelled as such. Numbers on the Outcomes page come from separate cohorts and must not be read as one combined series.
Peer-reviewed sources are preferred. Where a figure is a practice-level aggregate rather than a published study result, it is labelled as an aggregate.
Comparisons name their basis. A national or published comparison range is given with the literature it comes from, not asserted.
Nothing is published before its denominator is settled. Measures still being assembled appear on the Outcomes page marked as pending rather than estimated.
No outcome guarantees. Group results are not a prediction for an individual patient.

Measure definitions

90-day periprosthetic infection rate — approximately 0.3%

Counts: Periprosthetic joint infection identified within 90 days of the index procedure
Denominator: 13,593 same-day hip and knee replacements
Window: 90 days from date of surgery
Setting: TJS ambulatory surgery centers
Comparison basis: Approximately 1.0%, the commonly cited range in the general arthroplasty literature
Excludes: Superficial wound problems not meeting periprosthetic joint infection criteria; infections presenting after the 90-day window
Source: TJS same-day discharge arthroplasty series; see Publications cited

Same-day discharge, outpatient total hip replacement — 99.8%

Counts: Patients discharged to home on the day of surgery
Denominator: 587 patients in a published outpatient total hip replacement cohort
Window: Day of surgery
Setting: Freestanding ambulatory surgery center
Comparison basis: 30–48%, the published national range for same-day discharge after total hip replacement
Excludes: Patients not selected for an outpatient pathway; inpatient hospital cases
Important: This is a rate among patients already selected as candidates for outpatient surgery, not a rate among all hip replacement patients
Source: TJS outpatient total hip replacement series; see Publications cited

Readmission or unplanned intervention — 0.83%

Counts: Patients requiring hospital readmission or an unplanned intervention after discharge
Denominator: 600 patients in a published same-day discharge series (5 events)
Window: 90 days from date of surgery
Setting: Specialized ambulatory surgery center
Comparison basis: Approximately 2–5% in comparable published same-day discharge series
Important: This cohort of 600 is a different and smaller series than the 13,593-patient cohort used for the infection rate. The two figures describe different groups of patients.
Source: TJS same-day discharge arthroplasty series; see Publications cited

Same-day hip and knee replacement experience — 13,593

Counts: Primary total hip and total knee replacements performed on a same-day discharge pathway and included in published TJS analyses
Window: The study period of the underlying publication
Setting: TJS ambulatory surgery centers
Excludes: Inpatient cases; revision procedures; cases outside the published study period
Source: TJS same-day discharge arthroplasty series; see Publications cited

Combined hip and knee replacements — 100,000+

Counts: Completed hip and knee replacements across the career experience of the current TJS surgeon group
Update cadence: Recalculated weekly as a de-identified aggregate from the TJS clinic platform, adding completed cases to a stated practice baseline
Type: Practice-level aggregate, not a published study result
Excludes: Any patient-identifiable information; the aggregate returns totals only
Source: TJS clinic platform aggregate

Patient-reported outcome measures

TJS collects KOOS JR for knee replacement and HOOS JR for hip replacement. Both are validated, joint-specific instruments scored from 0 to 100, where higher scores indicate better joint function and less pain. The clinically meaningful unit is not the raw score but the change from before surgery to follow-up, and in particular whether a patient reaches the minimal clinically important difference (MCID) — the smallest improvement a patient actually notices.

When TJS PROM figures are published on the Outcomes page they will report, for each instrument: the number of patients with paired pre-operative and follow-up scores, the follow-up interval, the mean scores at each point, the proportion reaching MCID, and the proportion of eligible patients who actually returned a follow-up score. That last figure matters: a PROM result drawn from a small, self-selected subset of patients is not comparable to one drawn from most of a cohort, and a center reporting PROMs from 20% of its patients should not be read alongside one reporting from 80%.

TJS has published on the relationship between reaching the KOOS JR MCID at one year and patient satisfaction after total knee replacement (Guild et al., Arthroplasty Today 2025; PMID 40661697).

What public data can and cannot measure

A natural question about any practice's self-published outcomes is why they are needed at all, when the federal government already publishes hospital quality measures. TJS surgical leadership tested that directly, building the most complete arthroplasty composite public CMS data permits across every US hospital with a reported hip and knee complication and readmission score.

The analysis found that public reporting cannot support a judgment about arthroplasty quality or value, for reasons that follow from how the measures were designed rather than from how they are analyzed. The CMS patient-reported outcomes file publishes no result for any hospital in the country. CMS classifies 1,692 of 1,706 hospitals as no different from the national rate on the complication measure, so 99.2% cannot be told apart on the measure that would dominate any composite. Only 29.5% of listed hospitals have enough reported data to be scored at all, and those that do are systematically higher-volume. No arthroplasty-specific measure of cost is published.

These properties are not accidental. Risk-standardized complication and readmission measures exist to administer payment programs: they must adjudicate within a bounded period, attribute to contracted facilities rather than to surgeons or episodes, report on a multi-year lag, and avoid accusing an institution wrongly — which is exactly why almost every hospital is classified as average. Each choice is correct for a penalty program and a poor fit for a patient choosing where and with whom to have surgery.

The four-domain Arthroplasty Value Index previously described here is retired. It relied on a CJR data file that is not publicly distributed, included a patient-reported outcome domain that turns out to be uncomputable for every US hospital, and reported a convergent validity figure that reflected a same-period comparison against a measure already inside the index. No score was ever published for TJS or any other program, and none should be inferred. The current analysis is at /avi.

Review and correction

Figures on the Outcomes page are reviewed by TJS clinical leadership before publication and at least annually thereafter, under the practice medical editorial policy. A figure is revised or removed when its underlying study is superseded, when a denominator or follow-up period changes, or when a more current cohort is available.

If you believe a number on these pages is wrong or misleading, please tell us. Corrections are made on the page and reflected in the machine-readable dataset at /outcomes/tjs-outcomes.json.

Important: this page describes how group results were measured. It is educational and is not medical advice. Whether joint replacement is right for you, and what result you can expect, requires an evaluation that includes imaging, your medical history, and a surgeon's judgment.