Readmission Risk Score: LACE Index (0–19), HOSPITAL Score (0–13), CMS HRRP & EHR Prediction Models

Informational reference for hospital quality directors, health services researchers, and clinical decision support (CDS) informatics teams. This guide summarizes published readmission prediction models and CMS quality measurement methodologies and does not provide individual patient discharge instructions.

Hospital medicine investigators, care transition managers, and clinical informatics architects use this guide to calculate and compare the two most widely validated 30-day hospital readmission risk scores—the LACE Index (van Walraven et al., 2010) and the HOSPITAL Score (Donzé et al., 2013)—distinguish point-of-care risk stratification from the CMS Hospital Readmissions Reduction Program (HRRP) Excess Readmission Ratio (`ERR`), and audit EHR-embedded machine learning readmission models.

What Is a Readmission Risk Score? (Short Answer)

A readmission risk score is a validated clinical prediction model that estimates an inpatient’s probability of experiencing an unplanned hospital readmission (or death) within 30 days of hospital discharge. By combining administrative, demographic, comorbidity, and physiological variables available in the electronic health record (EHR) prior to discharge, models such as the LACE Index (`0–19` points) and the HOSPITAL Score (`0–13` points) stratify patients into low-, moderate-, and high-risk tiers so transitional care teams can target intensive post-discharge interventions—including pharmacist medication reconciliation, 7-day outpatient follow-up, and remote patient monitoring—to patients most likely to benefit.

1. The LACE Index (`0–19` Points, van Walraven et al., 2010)

Derived and externally validated by Carl van Walraven and colleagues in the Canadian Medical Association Journal (CMAJ) (2010) across 4,812 medical and surgical patients (and 1,000,000 administrative records in Ontario), the LACE Index predicts the composite risk of unplanned readmission or death within 30 days of discharge using four variables (`L`, `A`, `C`, `E`):

LACE Component	Clinical / Administrative Variable Level	Points Assigned
L — Length of Stay (days) (including day of admission and discharge)	< 1 day: 0 pts | 1 day: 1 pt | 2 days: 2 pts | 3 days: 3 pts | 4–6 days: 4 pts | 7–13 days: 5 pts | ≥ 14 days: 7 pts	0 to 7 points
A — Acuity of the Admission	Emergent or urgent admission (e.g., admitted via the Emergency Department): 3 pts Elective admission: 0 pts	0 or 3 points
C — Comorbidity Burden (Charlson Comorbidity Index score)	Charlson score 0: 0 pts | 1: 1 pt | 2: 2 pts | 3: 3 pts | ≥ 4: 5 pts (e.g., MI = 1, CHF = 1, COPD = 1, Diabetes w/o complications = 1, Renal disease = 2, Metastatic cancer = 6)	0 to 5 points
E — Emergency Department Visits (in the 6 months preceding the index admission, excluding the visit leading to the current admission)	0 ED visits: 0 pts | 1 visit: 1 pt | 2 visits: 2 pts | 3 visits: 3 pts | ≥ 4 visits: 4 pts	0 to 4 points

LACE Index Risk Stratification Thresholds

• Low Risk (`0–4` points): ~2.0% to 4.9% 30-day readmission or death rate.
• Moderate Risk (`5–9` points): ~6.5% to 12.2% 30-day readmission or death rate.
• High Risk (`≥ 10` points, up to `19` max): ≥ 14.3% to 43.7% 30-day unplanned readmission or death rate (overall model C-statistic = 0.684).
• LACE+ Extension (van Walraven et al., 2012): Adds patient age, sex, teaching hospital status, acute diagnoses/procedures during index admission, and alternative-level-of-care days (`-15` to `90` scale, C-statistic = `0.771`).

2. The HOSPITAL Score (`0–13` Points, Donzé et al., 2013)

Published by Jacques Donzé, Drahomir Aujesky, Derek Williams, and Jeffrey L. Schnipper in JAMA Internal Medicine (2013) and internationally validated across 117,065 patients in 9 hospitals across 4 countries (2016), the HOSPITAL Score specifically targets potentially avoidable 30-day medical readmissions by incorporating discharge laboratory values (`Hemoglobin` and `Sodium`) and inpatient service type:

HOSPITAL Acronym Letter	Clinical Predictor Criterion at Discharge	Points
H — Hemoglobin at discharge	Last measured hemoglobin prior to discharge < 12 g/dL (< 120 g/L)	+1 point
O — Oncology service discharge	Discharged from an Oncology clinical service (or active oncology treatment)	+2 points
S — Sodium level at discharge	Last measured serum sodium prior to discharge < 135 mEq/L (< 135 mmol/L)	+1 point
P — Procedure during index stay	At least one ICD-9-CM / ICD-10-PCS coded procedure performed during the hospitalization	+1 point
I — Index admission type	Urgent or emergent admission (non-elective)	+1 point
T — Admissions during past 12 months	0–1 admissions: 0 pts | 2–5 admissions: +2 pts | > 5 admissions: +5 pts	0, +2, or +5 points
AL — Length of Stay ≥ 5 days	Hospital length of stay ≥ 5 days	+2 points

HOSPITAL Score Stratification: Total scores (`0–13` points) stratify medical discharges into Low Risk (`0–4` points, ~5.8% potentially avoidable readmission), Intermediate Risk (`5–6` points, ~11.9%), and High Risk (`≥ 7` points, ~22.8% potentially avoidable readmission; C-statistic `0.69–0.72`).

How Is a Readmission Risk Score Used in Clinical Research and Healthcare Workflows?

In hospital quality operations and clinical informatics, readmission prediction operates across three distinct layers:

1. Real-Time EHR Discharge Workflow Integration: EHR systems (such as Epic Discharge Readiness views and Oracle Health Care Management) automatically compute the LACE Index, HOSPITAL Score, or proprietary gradient-boosted / logistic machine learning readmission scores daily using real-time ADT length of stay, active ICD-10-CM problem lists, and laboratory results. Because final billing codes are often assigned post-discharge, EHRs must validate predictive performance using real-time problem-list data under internal vs. external validation and TRIPOD+AI standards.
2. Transitional Care Intervention Allocation (Preventing Hospital Readmissions): High-risk patients (`LACE ≥ 10` or `HOSPITAL ≥ 7`) trigger automated care management pathways: bedside pharmacist teach-back, post-discharge Transitional Care Management (`CPT 99495` with 7-day interactive contact and 14-day face-to-face visit, or `CPT 99496` with 7-day face-to-face visit), and hospital-at-home and remote patient monitoring enrollment.
3. CMS Hospital Readmissions Reduction Program (HRRP, Social Security Act § 1886(q)): Unlike point-of-care LACE/HOSPITAL scores, CMS HRRP evaluates hospital-wide financial penalties (up to 3% of base inpatient DRG payments) using retrospective, claims-based hierarchical logistic regression models that compute each hospital’s Observed-to-Expected (O/E) Excess Readmission Ratio (ERR) across 6 clinical cohorts: Acute Myocardial Infarction (AMI), Heart Failure (HF), Pneumonia, Chronic Obstructive Pulmonary Disease (COPD), Coronary Artery Bypass Graft (CABG), and Elective Primary Total Hip/Knee Arthroplasty (THA/TKA), tracked alongside hospital core measures and eCQMs.

Frequently Asked Questions

What is readmission risk score?

A readmission risk score—most commonly the LACE Index (`0–19` points, van Walraven et al., 2010) or the HOSPITAL Score (`0–13` points, Donzé et al., 2013)—is a clinical prediction algorithm that combines hospital length of stay, admission acuity, comorbidity burden, prior emergency/inpatient utilization, and discharge laboratory values to stratify a patient’s risk of unplanned 30-day hospital readmission.

How is readmission risk score used in clinical research and healthcare?

Hospitals embed readmission risk scores into EHR discharge planning dashboards to identify high-risk patients (`LACE ≥ 10` or `HOSPITAL ≥ 7`) prior to discharge, trigger Transitional Care Management (`CPT 99495/99496`) and pharmacist medication reconciliation workflows, and reduce avoidable readmissions measured under the CMS Hospital Readmissions Reduction Program (HRRP).

Sources

• van Walraven C, Dhalla IA, Bell C, et al. Derivation and validation of an index to predict early death or unplanned readmission after discharge from hospital to the community. CMAJ. 2010;182(6):551–557. https://pmc.ncbi.nlm.nih.gov/articles/PMC2845681/ (2026-11-09).
• Donzé J, Aujesky D, Williams D, Schnipper JL. Potentially avoidable 30-day hospital readmissions in medical patients: derivation and validation of a prediction model. JAMA Internal Medicine. 2013;173(8):632–638. https://pmc.ncbi.nlm.nih.gov/articles/PMC3740920/ (2026-11-09).
• Centers for Medicare & Medicaid Services (CMS). Hospital Readmissions Reduction Program (HRRP). https://www.cms.gov/medicare/quality/value-based-programs/hospital-readmissions (2026-11-09).