8.2 Survey Sampling, Methodologies & Digital CAHPS Administration

8.2 Survey Sampling, Methodologies & Digital CAHPS Administration

Quick Answer: HCAHPS sampling requires identifying eligible adult inpatients (age 18+ with at least one overnight stay, alive at discharge, and a non-psychiatric primary MS-DRG). Surveys must be administered via continuous random sampling between 48 hours and 42 calendar days (6 weeks) post-discharge. To ensure statistical reliability on CMS Care Compare, hospitals target 300 completed surveys annually. Approved administration modes include Mail-only, Telephone-only, Mixed mode, Active IVR, and Web-first/Digital CAHPS. CMS applies Patient Mix Adjustment (PMA) regression to eliminate demographic and survey-mode bias.

Rigorous sampling and administration standards ensure that HCAHPS data reflect authentic patient experiences rather than sampling artifacts, mode biases, or unadjusted patient demographics. Experience leaders must understand these technical parameters to oversee survey vendors and interpret operational performance.

---

Patient Eligibility & Exclusion Standards

To ensure consistent cross-hospital comparisons, CMS establishes precise, non-negotiable criteria governing which discharged patients can be included in the survey sampling frame:

+-----------------------------------------------------------------------------------+
|                         HCAHPS PATIENT ELIGIBILITY CRITERIA                       |
+-----------------------------------------------------------------------------------+
|  1. AGE REQUIREMENT: Must be 18 years of age or older at the time of admission.   |
|  2. ADMISSION STATUS: Formal inpatient admission with at least ONE overnight stay.|
|  3. DISCHARGE STATUS: Discharged alive to home or community setting.              |
|  4. CLINICAL SERVICE LINE: Non-psychiatric primary diagnosis upon discharge       |
|     (Eligible service lines: Medical, Surgical, or Obstetric / Maternity).        |
+-----------------------------------------------------------------------------------+

Mandatory Patient Exclusion Categories

CMS strictly prohibits surveying certain categories of patients due to clinical vulnerability, legal constraints, or data integrity protocols:

---

Sampling Protocols & Administration Timelines

CMS enforces strict rules governing how patient records are sampled and when survey outreach must occur:

                    HCAHPS SURVEY ADMINISTRATION TIMELINE

  [ PATIENT DISCHARGE ]
          |
          |  48-HOUR MANDATORY RECOVERY BUFFER (No survey contact permitted)
          v
  [ EARLIEST CONTACT: DAY 3 (48 hrs post-discharge) ]
          |
          |  SURVEY OUTREACH WINDOW (Mail, Phone, Web, IVR)
          |
          v
  [ HARD CLOSE: DAY 42 (6 weeks post-discharge) ]
  (All survey attempts must cease; unreturned surveys marked as non-response)

1. Continuous Random Sampling (CRS) vs. Census

• Continuous Random Sampling (CRS): Hospitals with large discharge volumes draw a random sample of eligible patients on a continuous, monthly basis across all eligible service lines (Medical, Surgical, Maternity).
• Complete Census: Hospitals with smaller discharge volumes (fewer than 100 eligible discharges per month) are required to survey 100% of all eligible discharges (a census) to accumulate enough completed surveys for statistical validity.

2. The 300 Completed Survey Target

• CMS requires hospitals to target a minimum of 300 completed HCAHPS surveys over four consecutive rolling quarters (an average of 75 completed surveys per quarter).
• This sample size ensures a 95% confidence interval with a margin of error of approximately +/- 5.6 percentage points at the hospital level, providing sufficient statistical power for public Care Compare Star Ratings and HVBP payment adjustments.
• For very small rural or Critical Access Hospitals (CAHs) unable to reach 300 completed surveys despite a 100% census, CMS publishes their scores with an explanatory footnote indicating smaller sample size.

3. Survey Administration Cadence (48 Hours to 42 Days)

• Earliest Permissible Contact: Outreach cannot begin until 48 hours (2 full calendar days) following discharge. This ensures the patient has arrived home, rested, and separated from acute discharge stress.
• Latest Permissible Completion: All survey data collection must close within 42 calendar days (6 weeks) post-discharge. Any survey received or interview completed after Day 42 must be discarded and cannot be submitted to CMS.

---

Approved Survey Administration Modes & Modern Digital CAHPS

To accommodate diverse patient demographics, CMS approves five distinct survey administration modes. Each mode follows strict operational protocols to prevent bias:

                          APPROVED HCAHPS ADMINISTRATION MODES

   1. MAIL-ONLY          2. TELEPHONE-ONLY       3. MIXED MODE           4. ACTIVE IVR
  [ First Survey Mail ]  [ CATI Live Interview]  [ Mail Survey First ]  [ Automated Call ]
  [ 2nd Mail / Postcard] [ Up to 5 Attempts   ]  [ Phone Follow-up   ]  [ Voice/Keypad   ]
  
                       5. WEB-FIRST / DIGITAL CAHPS (Modernized)
                      [ Secure Email / SMS Digital Invite ]
                                        |
                               (If not completed)
                                        v
                      [ Follow-up via Mail or Telephone ]

Detailed Analysis of Administration Modes

Exam Tip: On the CPXP exam, remember that hospitals cannot choose an unapproved mode (such as point-of-care tablet surveying in the hospital discharge lounge) for official HCAHPS reporting. Point-of-care surveys violate CMS protocol because the patient is still on-site, introducing profound social desirability and coercion bias.

---

Patient Mix Adjustment (PMA) & Risk Adjustment Methodology

A critical requirement of fair public reporting is that a hospital treating an older, sicker, or socioeconomically disadvantaged population should not be penalized compared to a specialty hospital treating younger, healthier elective surgery patients. To eliminate non-hospital confounding factors, CMS applies Patient Mix Adjustment (PMA) to raw HCAHPS scores.

+-----------------------------------------------------------------------------------+
|                    PATIENT MIX ADJUSTMENT (PMA) REGRESSION MODEL                  |
+-----------------------------------------------------------------------------------+
|  RAW HOSPITAL SURVEY DATA                                                         |
|  - Unadjusted patient responses across all survey items                           |
|                                        |                                          |
|                                        v                                          |
|  MULTIVARIABLE OLS REGRESSION MODEL ADJUSTS FOR:                                  |
|  1. Patient Age (categorized into standardized age brackets)                      |
|  2. Educational Attainment (8th grade or less to 4-year college degree+)         |
|  3. Self-Rated General Health Status (Excellent, Very Good, Good, Fair, Poor)     |
|  4. Self-Rated Mental Health Status (Excellent, Very Good, Good, Fair, Poor)      |
|  5. Clinical Service Line (Medical vs. Surgical vs. Obstetric / Maternity)        |
|  6. Language Spoken at Home (English, Spanish, Chinese, Vietnamese, Other)        |
|  7. Survey Administration Mode (Mail, Phone, Mixed, IVR, Web-First)               |
|  8. Response Lag Time (Number of days elapsed between discharge and response)     |
|                                        |                                          |
|                                        v                                          |
|  STANDARDIZED, RISK-ADJUSTED HCAHPS TOP-BOX SCORES                                |
|  - Level-set playing field published on Care Compare & utilized in HVBP scoring   |
+-----------------------------------------------------------------------------------+

Why Specific Patient Mix Factors Matter

• Self-Rated Health Status: Empirical research demonstrates that patients who rate their overall health as "Poor" or "Fair" consistently rate their care lower than patients in "Excellent" health, regardless of hospital quality. PMA adjusts scores upward for hospitals treating sicker patient cohorts.
• Educational Attainment: Patients with higher formal education (college graduates) tend to be more critical of communication and rate hospital experiences slightly lower than patients with less formal education. PMA adjusts for local educational demographics.
• Service Line Differences: Maternity (obstetric) patients generally award significantly higher top-box ratings than medical inpatients suffering from chronic multi-morbid conditions. PMA balances differences in hospital clinical service mix.
• Mode Adjustment: Patients responding via telephone interviews historically award top-box ratings 3 to 5 percentage points higher than patients responding via mail (due to social desirability bias when speaking to a live human). PMA levels the playing field across different survey modes.

---

Vendor Certification & Survey Integrity Rules

CMS enforces strict guidelines to preserve the scientific integrity of HCAHPS administration:

1. Approved Survey Vendors: Hospitals cannot self-administer HCAHPS surveys unless they undergo rigorous CMS training and obtain formal authorization as a self-administering hospital. The vast majority of hospitals contract with CMS-certified third-party survey vendors (e.g., Press Ganey, Qualtrics).
2. Strict Prohibition of Coaching and Influencing:
   • Hospital staff are strictly prohibited from attempting to influence patient responses.
   • Prohibited Actions: Asking patients to give "all 10s" or "Always" ratings, showing copies of the HCAHPS survey to patients prior to discharge, offering incentives or gifts tied to survey completion, or telling patients that hospital funding depends on their survey answers.
   • Permissible Actions: Informing patients that they may receive a survey in the mail or by phone, emphasizing that their honest feedback is valued to improve care, and providing high-quality care during the entire stay.