HCAHPS Survey Administration: Timing, Modes, and the Fixed Question Set

HCAHPS (the Hospital Consumer Assessment of Healthcare Providers and Systems survey) is the national, standardized patient-experience survey every acute-care hospital participating in Medicare’s Hospital Inpatient Quality Reporting (IQR) Program must administer. For a patient-safety officer, quality director, or risk manager, the operationally important part usually isn’t the survey’s theory of patient experience — it’s the protocol: exactly when a discharge becomes eligible to be surveyed, which administration modes CMS actually allows, which questions can and cannot be changed, and how the resulting scores flow into Hospital Value-Based Purchasing (VBP) payment. This page works through that protocol directly.

Developed by the Agency for Healthcare Research and Quality (AHRQ) in partnership with the Centers for Medicare & Medicaid Services (CMS), HCAHPS has been the mandatory national patient-experience-of-care instrument since 2006, with public reporting required under the Hospital IQR Program since 2008. It is administered under CMS’s HCAHPS Quality Assurance Guidelines, which specify the eligible population, the timing window, the approved modes, and the fixed question wording — a hospital does not have latitude to redesign any of these.

What HCAHPS measures, and what it deliberately doesn’t

HCAHPS is a patient-experience survey, not a patient-safety-event survey. It asks discharged patients to rate specific, observable aspects of their care — whether nurses explained things clearly, whether pain was addressed, whether discharge instructions were understandable — not whether a harm event occurred. It is a distinct instrument from CAHPS surveys built for other settings (Clinician & Group CAHPS, Home Health CAHPS, and others in the same CAHPS family), and distinct from clinical patient-safety measures like AHRQ’s Patient Safety Indicators or NHSN healthcare-associated infection surveillance, which use administrative and surveillance data rather than patient self-report. Treating a low HCAHPS composite as evidence of a safety event, or a strong one as evidence of safety, conflates two different measurement systems that happen to sit on the same CMS reporting calendar.

Who actually gets surveyed: the eligible discharge population

HCAHPS is not sent to every patient who passes through a hospital. The eligible population is adult (18 and older) inpatients discharged alive, drawn continuously from medical, surgical, and maternity care service lines. CMS’s Quality Assurance Guidelines exclude specific discharge categories from the sampling frame — psychiatric MS-DRGs, patients discharged to hospice, prisoners, and non-U.S. residents are among the standard exclusions — and hospitals must apply the exclusion logic exactly as specified rather than substituting their own judgment about which discharges are “appropriate” to survey. A hospital (or its survey vendor) samples continuously, month over month, rather than surveying in occasional batches; CMS’s public reporting is built on four rolling quarters of continuously collected data, not a single snapshot period.

The administration window: 48 hours to six weeks after discharge

This is the single most operationally consequential rule on this page. HCAHPS surveys must be administered no earlier than 48 hours after discharge and generally no later than six weeks (42 days) after discharge. The floor exists so patients aren’t surveyed before they’ve had time to reflect on the stay (and, for mail administration, before a mailing sent immediately at discharge would even plausibly arrive); the ceiling exists because patient recall of a specific inpatient stay degrades sharply past that point, and CMS’s comparability model assumes a roughly consistent recall window across every participating hospital.

For a mixed-mode or telephone protocol with a follow-up wave, this window governs the entire administration sequence, not just the first contact — initial mailing or first call attempt, follow-up mailing or call attempts, and the point at which a case is closed out as a non-response all have to land inside the 48-hour-to-six-week frame. A vendor or in-house team that starts the clock from admission, from the date of a billing close, or from any date other than the actual discharge date is running a non-compliant protocol regardless of how well the rest of the process is executed.

Approved administration modes

CMS’s Quality Assurance Guidelines specify a fixed set of approved administration protocols; a hospital cannot invent its own delivery mechanism (e.g., a bedside tablet survey, a patient-portal pop-up, or a discharge-desk handout) and count it as HCAHPS. The core protocols are:

• Mail only — a paper survey mailed to the patient’s address of record, with a defined follow-up mailing to non-respondents.
• Telephone only — live interviewer-administered calls, following a scripted protocol and call-attempt schedule.
• Mixed mode — an initial mail survey, with telephone follow-up of patients who don’t return the mailed survey within the specified window.
• Active Interactive Voice Response (IVR) — an automated outbound telephone survey using recorded prompts and touch-tone or spoken response capture, rather than a live interviewer.

CMS’s current HCAHPS Survey Instruments materials also list web-based survey materials alongside the mail and phone versions, reflecting web/email administration options added to the approved protocol set in more recent Quality Assurance Guidelines revisions — hospitals evaluating a vendor’s proposed protocol should confirm against the current-year Quality Assurance Guidelines (published annually by CMS/HCAHPS) exactly which combination that vendor is approved to run, rather than assuming a protocol used in a prior survey year is still the current option set. Whichever mode is used, a hospital either contracts with a CMS-approved HCAHPS survey vendor from the published approved-vendor list, or self-administers under a separate CMS approval process with its own oversight requirements — there is no unsupervised, unapproved third option.

The fixed question set: what’s actually on the survey

The HCAHPS instrument is fixed wording, fixed response scales, and a fixed question order — a hospital or vendor cannot reword an item, add a Likert point, or reorder sections and still call the output HCAHPS. Hospitals are permitted to append a limited number of their own supplemental questions after the core instrument, but those supplemental items are never part of the scored, publicly reported measures.

The scored core of the survey is built from 15 individual survey questions that combine into six HCAHPS composite measures: Communication with Nurses, Communication with Doctors, Responsiveness of Hospital Staff, Communication about Medicines, Discharge Information, and Care Transition. Two further single-item measures — Cleanliness of the Hospital Environment and Quietness of the Hospital Environment — are reported individually rather than combined into a composite, and two global items — the 0–10 Overall Hospital Rating and the Recommend the Hospital item — round out the set. Together, these are the ten HCAHPS measures CMS publicly reports on Care Compare. Beyond that scored core, the full mail instrument runs longer once screening/skip-pattern items and the unscored “About You” demographic section (self-reported health status, education level, and similar background items, used for patient-mix adjustment rather than for scoring) are included.

Patient-mix adjustment matters here specifically because it’s the mechanism that makes cross-hospital comparison defensible at all: a hospital with a sicker, older, or less-English-proficient patient population is not compared raw against one with a healthier population. CMS applies a patient-mix adjustment model to each composite and item before public reporting, using factors including self-reported health status, education, service line, and a handful of other survey-collected variables — which is part of why the “About You” section exists on an instrument that otherwise looks purely experience-focused.

Sample size: two different thresholds, for two different purposes

Hospitals track HCAHPS sample size against two distinct CMS thresholds that are frequently conflated:

• The Hospital VBP inclusion floor — 100 completed surveys. Under 42 CFR 412.165(a)(1), Table 1, a hospital needs at least 100 completed HCAHPS surveys in the applicable period for HCAHPS to be scored at all in that year’s Value-Based Purchasing Total Performance Score. Fall short and the Person and Community Engagement domain isn’t calculated for that hospital — the domain drops out of the TPS entirely rather than being scored on thin data.
• The Star Ratings reliability threshold — 300 completed surveys. Separately from VBP eligibility, CMS’s HCAHPS Star Ratings methodology treats 300 completed surveys across the four rolling reporting quarters as the volume needed for a statistically reliable star rating; hospitals below that volume can have unreliable or suppressed star displays on Care Compare even though their underlying composite scores are still calculated and publicly reported as numbers.

A hospital can clear the 100-survey VBP floor comfortably and still land below the 300-survey Star Ratings reliability threshold — the two numbers answer different questions (does this measure count toward my payment adjustment, versus is my star display statistically dependable) and a quality team tracking only one of them is missing half the picture.

How the fixed HCAHPS scores become VBP dollars

HCAHPS composites and items don’t sit in isolation — they’re the entire input to the Person and Community Engagement domain of Hospital VBP, which converts into a real payment adjustment on every base operating DRG payment for the year. The administration protocol on this page determines whether a valid HCAHPS score exists at all; what happens to that score inside VBP — the achievement/improvement point arithmetic, domain weighting, the linear exchange function, and the FY 2027–FY 2029 modification that narrowed VBP scoring to six unchanged HCAHPS dimensions while the survey itself is being revised — is a full topic in its own right. CASRAI’s Total Performance Score guide works that arithmetic end to end with CMS’s own published figures; this page is the upstream half of that story — get the administration protocol wrong, and there’s no valid score for that later math to act on.

Operational checklist for patient-safety and quality teams

• Confirm the vendor’s protocol against the current-year Quality Assurance Guidelines, not last year’s contract terms — approved modes and supplemental-question limits are revised on a regular cycle.
• Audit the discharge-to-first-contact interval directly against the 48-hour floor; a vendor contacting patients too early is a compliance finding, not just a response-rate problem.
• Track sample size against both thresholds separately — the 100-survey VBP floor and the 300-survey Star Ratings reliability threshold — rather than one blended “are we surveying enough” number.
• Route dimension-level data to the unit, not just the composite score to the board. A hospital-level Communication with Nurses composite can mask a single unit dragging the whole score down; the underlying item-level data supports that drill-down even though only the composite is publicly reported.
• Don’t let HCAHPS become the safety programme. It’s one input to one VBP domain, not a surveillance system for harm — pair it with the hospital’s actual patient-safety event reporting rather than substituting for it.

Frequently asked questions

Can a hospital write its own HCAHPS questions?

Not for the scored instrument. The core question wording, order, and response scales are fixed by CMS. Hospitals may append a limited set of their own supplemental questions after the core survey, but those additions are never part of the publicly reported HCAHPS measures.

Does every discharged patient get an HCAHPS survey?

No. Only adult discharges from medical, surgical, and maternity service lines that meet CMS’s eligibility criteria enter the sampling frame; psychiatric discharges, hospice discharges, and several other categories defined in the Quality Assurance Guidelines are excluded.

What happens if a hospital doesn’t reach the minimum completed-survey count?

Below 100 completed surveys, HCAHPS drops out of that year’s VBP Total Performance Score entirely for that hospital. Below 300 completed surveys (a separate threshold), the hospital’s public-facing HCAHPS Star Rating can be flagged as statistically unreliable or suppressed, even though the underlying composite scores are still reported.

Is a bedside tablet or patient-portal survey the same as HCAHPS?

No. Those tools can be useful supplementary listening mechanisms, but they are not an approved HCAHPS administration mode and their results are not HCAHPS data. Official HCAHPS administration has to run through mail, telephone, mixed mode, active IVR, or the web-based option CMS has approved, via a CMS-approved vendor or an approved self-administration arrangement.

Does a bad HCAHPS score mean a patient-safety event occurred?

Not directly. HCAHPS measures patient-reported experience of care, not confirmed safety events. A hospital can score poorly on communication items with no safety event present, and a safety event can occur without moving HCAHPS scores at all — they are separate measurement systems that happen to both roll up into CMS quality reporting.