Inpatient care experiences are better in critical access hospitals than in other hospitals

Abstract

Prior studies have found that critical access hospitals (CAHs), which serve patients who would otherwise have limited access to hospitals, provide lower-quality clinical care than inpatient prospective payment system (IPPS) hospitals; evidence is limited about the patient experiences they provide. Using linear mixed-effects regression models, we compared patient-mix–adjusted Hospital Consumer Assessment of Hospitals, Providers, and Systems (HCAHPS) survey scores for CAHs and IPPS hospitals and evaluated how much of the observed differences were associated with size, location, and other hospital characteristics. CAH patients were older, more often in the medical service line, had lower educational attainment, and worse self-rated health than their IPPS counterparts. Accounting for such differences, CAH patients had better experiences (+8 points on the 0–100 HCAHPS summary score, where differences >5 are considered large by patient experience heuristics), especially for staff responsiveness, cleanliness, quietness, and discharge information. CAHs do not outperform similarly small IPPS hospitals, which often have different missions (eg, for-profit surgical specialty hospitals). For-profit and teaching status, while uncommon among CAHs, predicted lower CAH HCAHPS performance. Despite the limited services provided by CAHs, their small scale may facilitate positive experiences for patients in areas with limited hospital choices. For-profit and teaching CAHs may benefit from quality-improvement efforts.

Introduction

Critical access hospitals (CAHs) play an important role in rural populations' health care safety net by serving patients who otherwise may have limited or no access to hospitals.¹ By definition, CAHs have fewer than 25 beds (excluding up to 10 beds each for psychiatric and rehabilitation units, ambulatory care beds, beds used exclusively for obstetric delivery, and newborn bassinets and observation beds),² are located at least 35 miles from other hospitals, maintain annual average length of stay of 96 hours or less for acute-care-stay patients, provide 24/7 emergency care services, and are in rural areas or treated as rural under a special provision that qualifies hospitals in urban areas.³ As of 2025, 8 states had at least 50 CAHs (Illinois, Iowa, Kansas, Minnesota, Montana, Nebraska, Texas, Wisconsin) and 5 eastern states (Connecticut, Delaware, Maryland, New Jersey, and Rhode Island) and the District of Columbia had none (historical CAH data; Flex Monitoring Team).⁴

Critical access hospitals receive certain financial benefits, including cost-based reimbursement for Medicare (and in some states, Medicaid) services, flexible staffing, and services that may strengthen the hospital financially.¹ As part of the Medicare Rural Hospital Flexibility grant program, CAHs participate in a quality-improvement program for rural hospitals. In some states, CAHs may receive funding from the Health Resources and Services Administration to hire survey vendors to administer the survey.⁵

Despite this support, CAHs face multiple ongoing challenges, including chronic and worsening workforce shortages, higher labor costs, and recruiting difficulties.^6,7 For some CAHs, it is thought that public and private reimbursements are too little to cover fixed costs.⁸

Previous studies have found that the clinical quality of care for common medical admissions in CAHs compares unfavorably to other hospitals,^9,10 although the evidence is less clear for surgical admissions.^11,12 Building upon this prior work, in this study we used cross-sectional data from US hospitals to describe differences in patient experiences between CAHs and inpatient prospective payment system (IPPS) hospitals. Prior work shows that smaller hospitals provide better patient experiences for most domains than larger hospitals.¹³ One study further suggests that CAHs performed somewhat better than other hospitals on a composite measure of patient experience in 2008, a lead that had widened slightly by 2011.¹⁴ A second study found that CAHs outperformed others on communication about medications during 2013–2014.¹⁵ Neither study examined the role of other hospital characteristics, including size, in explaining the better patient experiences in CAHs. We predicted that patient experiences would be better in CAHs than IPPS hospitals and that much of this difference would be accounted for by the characteristics of CAHs and smaller size, which may be associated with a stronger focus on community needs and personalized care. We also assessed whether other geographic differences, such as local health care practice patterns, could explain differences in patient experiences reported by patients treated in CAHs.

Data and methods

Data

Hospital Consumer Assessment of Hospitals, Providers, and Systems (HCAHPS) surveys from 2 226 315 patients in 2022 from 4424 hospitals (24.1% response rate) were analyzed. The HCAHPS survey is administered to random samples of adult inpatients in the medical, surgical, and maternity service lines with an overnight stay. The survey is initiated between 48 hours and 42 days after discharge using 1 of 4 approved modes of administration: mail-only, telephone-only, mixed mode (mail with telephone follow-up of nonrespondents), and active interactive voice response.

Measures

The study's primary outcome was an HCAHPS summary score (HCAHPS-SS), the average of 10 HCAHPS measures: 6 multi-item composite measures that were each given a weight of 1.0 (Nurse Communication, Doctor Communication, Staff Responsiveness, Communication about Medicines, Care Transition, and Discharge Information). Two global items (Overall Hospital Rating, Hospital Recommendation) measure 1 construct together, so were each given a weight of 0.5, and 2 single items (Cleanliness and Quietness) were also each given a weight of 0.5. Survey response options were “never,” “sometimes,” “usually,” or “always” for all measures other than Discharge Information (options were “yes” or “no”), Care Transition (options were “strongly disagree,” “disagree,” “agree,” or “strongly agree”), Hospital Recommendation (options were “definitely no,” “probably no,” “probably yes,” or “definitely yes”), and Overall Hospital Rating (response options ranged from 0 [“worst possible hospital”] to 10 [“best possible hospital”]). Top-box scoring, the proportion of most-positive category responses (except for the overall hospital rating, for which 9 or 10 corresponds to a top-box response), was used for all HCAHPS measures and the HCAHPS-SS. Secondary analyses examined each of the 10 constituent HCAHPS measures individually. The HCAHPS items and details of survey implementation protocols can be found at www.hcahpsonline.org.

The HCAHPS survey also asks respondents about their age, education, self-rated overall health, self-rated mental health, and language spoken at home (Chinese, English, Spanish, other).

Other measures

Prior research has shown that hospital characteristics, including teaching status, ownership status, system affiliation, rural/urban location, bed size, and nurse staffing levels,^16-20 are associated with HCAHPS scores.¹³ We used 2022 data from the American Hospital Association (AHA) to assess each of these variables, with 1 exception: CAH bed size was taken from the Flex Monitoring Team (Critical Access Hospital Locations List; Flex Monitoring Team²¹), which provides more accurate counts for CAHs. Staffing level was defined as the number of full-time–equivalent nurses, registered nurses, and licensed practical nurses working in the hospital (excluding nursing home and long-term care unit staff) per 1000 adjusted patient-days. We used a mean imputation for missing data (<1%) on nursing levels. Staffing levels were then classified into quartiles. We used the 2022 AHA data to classify each hospital as a CAH or IPPS hospital.

Analyses

We first compared CAHs with all IPPS hospitals and with very small (<30 beds) IPPS hospitals with respect to the following characteristics: rural/urban location (urban being a location inside a Metropolitan Statistical Area [MSA] and rural being a location outside an MSA), teaching status (major teaching hospital, minor teaching hospital, nonteaching), ownership status (government, nonprofit, for-profit), system affiliation, and bed size), state/district/territory, and number of completed HCAHPS surveys. We also compared the patient characteristics, the official HCAHPS patient-mix adjusters (age, education, self-rated overall health, self-rated mental health, language spoken at home [Chinese, English, Spanish, other], and sex-by-service line²²) from administrative records and the survey. Missing patient-mix adjusters were imputed as the hospital mean.

As detailed in Appendix 1, we estimated a series of 11 linear regression models (1 for HCAHPS-SS and each HCAHPS measure) predicting mode-adjusted HCAHPS measures from the CAH indicator, HCAHPS patient-mix adjustors (age, education, self-rated health, response percentile, language spoken at home [Chinese, English, Spanish, and other], and sex-by-service line interactions), and random hospital intercepts. Linear rather than logistic regression models are appropriate because they are almost identical when sample sizes are large and outcomes are predominantly between 20% and 80%²³ (as was the case here²⁴). To assess the contribution of different geographic and other hospital factors to observed differences, we estimated additional models that added fixed effects for health referral regions (HRRs) that characterize health care markets, rural location, teaching status, ownership, and staffing levels among all hospitals and restricting to CAHs and small IPPS hospitals. We used patient experience magnitude heuristics: differences of 1, 3, and 5 points on a 0–100 scale were considered small, medium, and large, respectively.²⁵ We also calculated the CAH effect size using hospital-level SDs. Finally, we used a mixed linear regression model predicting mode-adjusted HCAHPS-SS that included fixed effects for rural location, teaching status, ownership, system affiliation, and teaching status among all CAHs, patient-mix adjusters, and hospital random effects.

Results

As shown in Table 1, 4% of all HCAHPS respondents were in CAHs, which were 29% of hospitals in this study. Overall, compared with all IPPS hospitals, all CAHs are more often rural (80% vs 20% for IPPS hospitals), nonteaching (88% vs 41% for IPPS), government-affiliated (42% vs 15% for IPPS), and have fewer than 50 beds (83% vs 18% for IPPS). Critical access hospitals were less likely to be for-profit (4% vs 21% for IPPS) and had higher staffing levels (highest-quartile staffing: 47% vs 16% for IPPS). Compared with patients in IPPS hospitals, CAH patients were older (median age: 74 vs 71 years for IPPS), more often in the medical service line (79% vs 56% for IPPS), had lower educational attainment (50% without any college vs 41% for IPPS), and were in worse self-rated health (31% in fair/poor health vs 28% for IPPS).

Table 1.

Inpatient and hospital characteristics by hospital type and size, 2022 HCAHPS survey.

	IPPS hospitals (n = 3138), %	CAHs (n = 1238), %	IPPS <30 beds (n = 284), %
Hospital characteristics
Rurala	23.4	79.6	30.0
Urban	76.6	20.4	70.0
Teaching status
Major	7.6	0.0	0.4
Minor	51.4	11.9	17.4
Nonteaching	41.0	88.2	82.2
Ownership
Government	15.0	41.6	17.8
Nonprofit	63.7	54.3	34.3
For-profit	21.3	4.1	47.9
System-affiliated	74.5	46.5	49.2
Bed size
6–24	6.1	100.0	78.1
25–49	11.8	0.0	21.9
50–99	16.2	0.0	0.0
100–199	25.8	0.0	0.0
200–299	15.4	0.0	0.0
300–399	9.5	0.0	0.0
400–499	5.4	0.0	0.0
500 or more	9.8	0.0	0.0
Quartiles (Q) of nurse-to-patient staffing ratio
Q1 (lowest staffing; <7.57)	24.5	25.8	8.5
Q2 (7.57–10.58)	31.5	9.9	10.2
Q3 (10.59–16.52)	28.3	17.4	21.3
Q4 (highest staffing; <16.52)	15.7	46.9	60.0
Patient characteristics
No.	2 166 358	98 957	51 819
Age
18–24 y	1.8	1.8 b	1.2
25–34 y	7.4	6.5	4.5
35–44 y	5.7	3.2	3.8
45–54 y	6.3	4.0	6.5
55–64 y	15.5	12.5	17.1
65–74 y	27.8	26.7	32.7
75–84 y	25.2	29.2	25.5
85+ y	10.3	16.0	8.7
Education
Eighth grade	3.9	3.9	2.7
Some high school	6.7	7.5	5.5
High school graduate	30.0	38.9	31.0
Some college or 2-y degree	29.2	29.3b	31.3
Four-year college graduate	14.4	10.4	14.2
Graduate school	15.8	10.1	15.2
Overall health
Excellent	11.6	9.6	12.0
Very good	25.2	22.9	29.7
Good	35.6	36.1	36.7
Fair	21.9	24.4	17.5
Poor	5.8	7.0	4.1
Mental health
Excellent	25.8	21.0	27.5
Very good	34.1	34.4	36.1
Good	28.0	31.2	26.9
Fair	10.4	11.6	8.4
Poor	1.8	1.8b	1.2
Sex by service line
Maternity	9.4	8.4	5.1
Female—surgery	16.8	7.0	26.6
Female—medical	29.2	42.7	24.9
Male—surgical	17.7	5.6	22.2
Male—medical	27.0	36.3	21.3

All non-italicized entries in the CAH column differ from the IPPS column entry (P < .05) and non-italicized entries in the IPPS hospital column differ from the corresponding CAH column entry (P < .05). Source: Authors’ own analysis of 2022 HCAHPS and 2022 AHA data.

^aRural/urban, with urban a location that is inside a Metropolitan Statistical Area (MSA) and rural a location outside an MSA.

^bEntry does not differ significantly (P < .05) from the corresponding IPPS entry.

Abbreviations: AHA, American Hospital Association; CAH, critical access hospital; HCAHPS, Hospital Consumer Assessment of Hospitals, Providers, and Systems; IPPS, inpatient prospective payment system.

Next, we compared the characteristics of the smallest (<30 beds) CAHs (n = 979) and IPPS hospitals (n = 284). Small CAHs typically reported rural (non-MSA) locations on the AHA annual survey (81%; the remaining CAHs also met CAH program rural location requirements), usually (76%) had highest-quartile staffing levels, and almost always (96%) were either government-affiliated (42%) or nonprofit (54%). In contrast, small IPPS hospitals were typically urban (70%), often for-profit (48% vs 4% for small CAHs), and had almost half of their patients in the surgical service line (49% vs 12% for small CAHs).

The mean (median) number of completed HCAHPS surveys were 77 (56) for CAHs and 695 (774) for IPPS hospitals. Almost three-fourths of CAHs (73%) but only 1 in 10 (10%) IPPS hospitals had fewer than the 100 completed HCAHPS surveys annually needed for reliable HCAHPS measurement.²⁶

Explaining the difference in performance of CAHs and IPPS hospitals

Table 2 (model 1) shows the average difference in patient experiences between CAHs and IPPS hospitals across all measures. On average, CAHs had HCAHPS scores that were 7.8 percentage points (pp) higher than IPPS hospitals. By the usual criterion of 5 pp as a large difference, the differences were large for all but 1 measure (Discharge Information, +2.2 pp). The largest difference was for Staff Responsiveness (+11.4 pp). Next, we assessed the extent to which these differences related to where CAHs were located and to the hospital characteristics by adding HRR, rural location, teaching status, ownership, system affiliation, and staffing level to model 1. As shown in model 2, location and hospital characteristics were associated with 41% [(7.8−4.6)/7.8] of the CAH difference in HCAHPS-SS, but moderate-to-large differences remained for all measures but Discharge Information. Restricting to only IPPS hospitals with fewer than 30 beds (model 3), CAHs no longer provided better patient experiences. Instead, there was a moderate negative difference in 1 measure: CAHs had worse Quietness (−3.9 pp; P < .001) than IPPS hospitals with fewer than 30 beds.

Table 2.

Models predicting HCAHPS top-box scores from critical access hospital indicator, other hospital indicators, all hospitals and small (<100 Beds) and smaller (<30 Beds) hospitals (on 0−100 scale).

Model 1: Predicting HCAHPS outcomes from CAH indicator
HCAHPS-SS			Nurse			Doctor			Staff			RX Comm			Clean
Beta	SE	P	Beta	SE	P	Beta	SE	P	Beta	SE	P	Beta	SE	P	Beta	SE	P
7.8	0.2	***	5.9	0.3	***	5.4	0.2	**	11.4	0.4	***	6.1	0.3	***	8.7	0.4	***
Quiet			Discharge			CTM			Rating			Recommend
Beta	SE	P	Beta	SE	P	Beta	SE	P	Beta	SE	P	Beta	SE	P
8.1	0.4	***	2.2	0.2	***	6.1	0.3	***	9.0	0.4	***	9.3	0.4	***

Model 2: Model 1 plus HRR, rural indicator, teaching status, ownership, and staffing levels
HCAHPS-SS			Nurse			Doctor			Staff			RX Comm			Clean
Beta	SE	P	Beta	SE	P	Beta	SE	P	Beta	SE	P	Beta	SE	P	Beta	SE	P
4.7	0.2	***	4.1	0.2	***	3.8	0.2	***	7.7	0.4	***	4.1	0.3	***	6.7	0.4	***
Quiet			Discharge			CTM			Rating			Recommend
5.0	0.4	***	0.9	0.2	***	4.0	0.3	***	6.2	0.4	***	6.6	0.5	***

Model 3: Restricts Model 2 to hospitals with <100 beds
HCAHPS-SS			Nurse			Doctor			Staff			RX Comm			Clean
Beta	SE	P	Beta	SE	P	Beta	SE	P	Beta	SE	P	Beta	SE	P	Beta	SE	P
3.5	0.3	***	3.0	0.3	***	2.6	0.3	**	5.6	0.4	***	3.2	0.4	***	5.1	0.4	***
Quiet			Discharge			CTM			Rating			Recommend
Beta	SE	P	Beta	SE	P	Beta	SE	P	Beta	SE	P	Beta	SE	P
4.4	0.5	***	1.0	0.2	***	3.3	0.4	***	4.7	0.5	***	5.3	0.5	***

Model 4: Restricts Model 2 to hospitals with <30 beds
HCAHPS-SS			Nurse			Doctor			Staff			RX Comm			Clean
Beta	SE	P	Beta	SE	P	Beta	SE	P	Beta	SE	P	Beta	SE	P	Beta	SE	P
-0.2	0.5		-0.1	0.5		0.1	0.6		-0.3	0.8		0.1	0.7		0.9	0.8
Quiet			Discharge			CTM			Rating			Recommend
Beta	SE	P	Beta	SE	P	Beta	SE	P	Beta	SE	P	Beta	SE	P
-3.3	0.9	**	0.5	0.4		-0.3	0.7		-0.4	0.8		0.1	0.9

***P<0.001; **P<.01.

Predicting performance among performing CAHs

Table 3 shows the characteristics associated with HCAHPS-SS among CAHs. System affiliation and rural location were not associated with performance, but teaching status was associated with a −1.6 pp (P < .001) HCAHPS-SS and for-profit status with a −2.6 pp HCAHPS-SS (P = .002).

Table 3.

Predicting HCAHPS summary score among 1238 CAHs (linear regression model).

	Beta	SE	P
Teaching	−1.6	0.4	<.001
Rural	−0.2	0.4	.619
Government	0.2	0.3	.526
Profit	−2.6	0.8	.002
Nonprofit	Ref	Ref	Ref
System-affiliated	<0.1	0.3	.915

Source: Authors’ own analysis of 2022 HCAHPS and 2022 AHA data. Model includes a random hospital effect.

Abbreviations: AHA, American Hospital Association; CAH, critical access hospital; HCAHPS, Hospital Consumer Assessment of Hospitals, Providers, and Systems; Ref, reference.

Limitations

Our study has several potential limitations. First, response rates were modest, and nonresponse bias may have affected the findings. However, research on HCAHPS surveys has found little evidence of nonresponse bias after adjustment for patient mix.²⁷ Second, although we controlled for self-rated health and service line, the more limited services provided by CAHs may result in less medically complex patients. In an analysis of Medicare enrollees hospitalized for 1 of 4 common surgical conditions nationwide, CAHs operated on less complex patients than other hospitals and had higher rates of triaging high-risk patients to larger centers before operation.¹¹ Third, the AHA measure may overestimate bed size, as evidenced by nearly two-thirds of CAHs reporting more than 25 beds.

Discussion

We found that CAHs, on average, provided substantially better experiences than IPPS hospitals (+8 points on the 0–100 HCAHPS-SS, where differences <5 are considered large), with large differences for all HCAHPS measures but Discharge Information. The largest difference (>11 points) was for Staff Responsiveness. But almost all measures, both relational and nonrelational (eg, Cleanliness, Quietness, and Discharge Information) had large differences from IPPS hospitals.

In addition to confirming the higher average patient experience performance for CAHs than IPPS hospitals, we provide stronger evidence than previously shown in the literature that structural characteristics, especially their small size, may be key factors in their higher performance.

Critical access hospitals are uniformly small and rural and have government or private nonprofit ownership more often than IPPS hospitals, all factors associated with better patient experiences. It is challenging to compare CAHs and IPPS hospitals because they are very different. The CAH vs IPPS differences in location, teaching status, ownership status, and staffing levels were associated with approximately 40% of differences. But perhaps the key difference is size. Critical access hospitals did not outperform IPPS hospitals with fewer than 30 beds. These comparisons, while increasing comparability of size, may result in even greater contrasts in missions. Compared with CAHs, very small IPPS hospitals (<30 beds) are much more often for-profit (48% vs 4% for CAHs). The smallest IPPS hospitals include many surgical specialty hospitals, with very different missions than CAHs, as well as more predictable admissions and less variability in reason for hospitalizations, which allows for greater pre-planning. Small for-profit physician-owned specialty hospitals have relied substantially on elective surgeries,²⁸ whereas most patients admitted to CAHs are through the medical service line.

We also found variation in HCAHPS performance among CAHs—some consistent with findings across all hospitals, and some unique to CAHs.^13,29 Consistent with findings in primarily IPPS hospitals, the few for-profit CAHs performed less well on HCAHPS than other CAHs. Whereas few differences in HCAHPS scores have been associated with teaching status among predominantly IPPS hospitals, we found that teaching CAHs performed less well on HCAHPS than nonteaching counterparts, suggesting that teaching may pose additional challenges in a CAH environment. These results suggest that for-profit and teaching CAHs (12%) may especially benefit from quality-improvement efforts. Because of smaller patient volume, individual CAHs may have greater challenges than most IPPS hospitals in reliably measuring and monitoring patient experience, making it difficult to evaluate their quality-improvement efforts: 73% do not collect 100 completed surveys per year. Critical access hospitals may need to pool data over multiple years to assess and improve their HCAHPS performance.

Despite the limited services provided by CAHs and many other small hospitals, their small scale may facilitate positive experiences for patients in areas with limited hospital choices. These findings are in contrast to other studies that document that CAHs may provide a lower quality of clinical care, on average, than IPPS hospitals in general, although prior research did not ascertain whether this was true when comparing CAHs with similarly sized IPPS hospitals.^9-12

The benefits of very small hospitals like CAHs may be greatest in the face of severe staffing shortages, such as those that occurred during the COVID-19 pandemic. A recent study found that small rural hospitals, because of chronic staffing shortages, are characterized by adaptive, multidimensional staff roles so that staff are accustomed to wearing multiple operational hats. Moreover, such hospitals may be better able to create a familial, collegial staff that may carry over and provide more individualized care settings for patients.^9,30 While some of these policies would be difficult to replicate in IPPS settings, there may be lessons to be learned about how to redirect limited staff resources to meet patient needs. The patient experience advantage of CAHs is concentrated in nonprofit and government-owned CAHs; for-profit CAHs provide substantially poorer patient experiences than other CAHs.

Conclusion

This observational study found that CAHs provide better patient experiences than IPPS hospitals and that much of this advantage is associated with structural characteristics, especially smaller size and rural location, which each seem to be associated with better patient experiences. It is challenging to directly compare CAHs and very small IPPS hospitals (those with <30 beds) because of differences in patients and missions. As with IPPS hospitals, nonprofit and government-owned CAHs outperform for-profit CAHs and may serve as best-practice models for small private CAHs and small IPPS hospitals.

Supplementary material

Supplementary material is available at Health Affairs Scholar online.

Funding

This study was funded by the Centers for Medicare and Medicaid Services (CMS) contract GS-10F-0166R/HHSM-500-2017-00077G to the Health Services Advisory Group.

Data availability

Data is available from the Centers for Medicare & Medicaid Services via a Data Use Agreement.