INTRODUCTION
Global aging population has accelerated rapidly as life expectancy continues to rise worldwide; in 2024, global life expectancy at birth reached 73.3 years, an increase of 8.4 years since 1995 [1]. The number of people aged 60 years or older is projected to increase from 1 billion in 2020 to 1.4 billion in 2030 and 2.1 billion by 2050 [2]. In response, Organization for Economic Co-operation and Development (OECD) countries are expected to see long-term care (LTC) expenditures rise from about 1% of GDP in 2005 to between 2% and 4% by 2050 [3]. In developed countries, approximately 2%~5% of the older population resides in nursing homes, and the proportion of facility-based LTC is expanding in countries such as Japan, Germany, and Italy due to rapid population aging and the weakening of traditional family caregiving [3,4]. Similarly, in Korea, the number of LTC facilities increased rapidly from 4,871 in 2014 to 6,323 in 2024, while the number of facility residents grew from 168,924 to 286,016 during the same period [5,6].
Ensuring high-quality medical and nursing care in long-term care facilities (LTCFs) has become a critical policy issue in OECD countries due to the aging population and increasing demand for LTC services [7]. Pressure injuries and urinary tract infections (UTIs) are among the most frequently used measures of quality of care in LTC settings, with high levels of international standardization achieved through expert consensus. UTIs are highly prevalent in nursing home residents, with prevalence rates ranging from 0.6% to 21.8% and incidence between 0.3 and 0.8 cases per 1,000 resident care days [8]. UTI treatment represents 30%~50% of antibiotic use in LTC facilities, contributing to antibiotic resistance and adverse effects. The use of indwelling urinary catheters to manage chronic voiding dysfunction ranges from 7%~10% among LTC residents, increasing UTI risk by 2.6 times.
Pressure injuries remain a significant issue in nursing homes, with prevalence rates ranging from 5.7% to 25.9% in some developed countries [9,10]. In intensive care units, the incidence rate can reach 59% [11]. These conditions significantly deteriorate residents’ health status and are associated with increased 30-day mortality [12].
UTIs and pressure injuries in LTCFs result from complex interactions between individual characteristics and organizational attributes, requiring a theoretical framework based on Ecological Systems Theory and Institutional Theory [13,14]. According to Bronfenbrenner’s ecological systems theory, individual health outcomes are determined through interactions among microsystems (personal characteristics), mesosystems (facility environments), and exosystems (policies and regulations), while Institutional Theory explains that healthcare organizations operate within institutional environments shaped by regulative, normative, and cultural-cognitive pressures, systematically influencing individual patient outcomes. From this theoretical perspective, adverse events among nursing home residents cannot be explained solely by individual biological risk factors, but require consideration of contextual factors such as staffing level, quality management systems, partnership with affiliated medical institutions, and organizational culture that may influence outcomes independently of or in interaction with individual-level effects.
Nursing home residents share common care environments and organizational characteristics within the same facility, creating a hierarchical data structure where individual health outcomes are not independent. Multilevel analysis is a statistical method that appropriately accounts for the hierarchical structure where individuals are clustered within institutions, enabling simultaneous estimation of individual-level and institutional-level effects [15]. This approach separates compositional effects from contextual effects, allowing accurate identification of organizational factors amenable to policy intervention. By evaluating independent effects at the facility level while controlling individual characteristics, multilevel analysis provides scientific evidence for developing effective prevention strategies and quality improvement policies.
Individual-level factors significantly associated with adverse health outcomes include advanced age, impaired mobility, cognitive decline, sex differences, obesity, poor nutritional status, and the presence of indwelling urinary catheters [16-19]. Female residents face higher UTI risk due to anatomical characteristics, while males may experience greater pressure injury risk due to body weight and muscle mass differences [19]. At the facility level, staffing adequacy is consistently identified as an essential determinant, with increased certified nursing assistant levels being associated with reduced pressure injury incidence and overall staffing influencing UTI rates [16,18].
Comprehensive databases such as the National Health Insurance Service (NHIS) database offer valuable opportunities to examine occurrence patterns and associated individual- and institutional-level factors. Previous research on adverse events among South Korean nursing home residents has been limited to small, facility-specific samples [20-22], constraining the generalizability of findings. Understanding contributing factors is particularly critical given the high prevalence and considerable preventability of these adverse events. Large-scale database analysis provides distinct advantages by enabling examination of factors that can be directly addressed through targeted policy interventions.
Using the NHIS database, this study investigates UTI and pressure injury experiences among Korean nursing home residents and identifies significant individual- and institutional-level factors associated with these events. The analysis leverages administrative data variables to provide evidence for developing targeted prevention strategies and quality improvement initiatives.
METHODS
This study was exempted from the Institutional Review Board (IRB) review by the National Health Insurance Service IRB (IRB No. 연-2021-HR-01-035) due to its retrospective design and the use of de-identified secondary data, which poses minimal risk to participants. The requirement for informed consent was waived by the IRB.
1. Study Design
This study comprehensively investigated UTI and pressure injury experiences among Korean nursing home residents from January 1, 2016 to December 31, 2017 using NHIS database.
We implemented a multi-level analytical approach to systematically decompose and examine the independent contributions of both individual-level and institutional-level factors associated with these adverse health events. This hierarchical modeling strategy allowed us to account for the nested structure of residents within facilities and to distinguish between patient-specific risk factors and organizational characteristics that influence clinical outcomes. We applied two models for both pressure injury and UTI: one including only individual-level variables (Model 1) and another incorporating both individual- and organizational-level variables (Model 2).
This study was written in accordance with the Strengthening the Reporting of Observational Studies In Epidemiology guidelines (https://www.strobe-statement.org).
2. Data Source and Study Samples
This study employed data from the NHIS database, focusing on Long Term Care Insurance and National Health Insurance claims records. We obtained permission from NHIS to access the NHIS database directly. We constructed a customized database to support the study’s objectives [23]. This study was conducted using the resulting database.
Figure 1 illustrates the process by which the study sample was selected. From the initial 5,304 nursing homes, 4,242 facilities met the inclusion criteria, participating in LTC facility evaluation in 2018 and having LTC insurance claim records in December 2017. Among the 215,742 residents in these facilities, 122,312 met the continuous residency criteria and constituted the final sample for analysis.
The sample included residents who continuously resided in nursing homes for at least 3 months between January 1, 2016, and December 31, 2017. This 3-month timeframe aligns with the internationally recognized threshold that distinguishes short- and long-term stays. Residents were considered to have used nursing home services in a given month if they stayed there for at least 1 day during that month.
Individual characteristics were extracted from the latest LTC needs assessment data collected between January 2016 and December 2017. Experience data regarding UTIs and pressure injuries were identified from national health insurance claims records during the residency period. The national health insurance claims records of nursing home residents were reviewed and confirmed by matching records corresponding to their period of nursing home residency.
3. Measurements
The primary outcomes were experience with pressure injuries (KCD-7 codes of L89, L890~L893) and UTIs (KCD-7 codes of N10~N13, N15, N16, N30, N34, N390), identified by the presence of these conditions as the primary diagnosis or among the first five secondary diagnoses in the nursing home residents’ NHI claims records.
We selected variables measurable within the NHIS administrative dataset and categorized them into individual- and institutional-level factors drawing on prior multilevel analyses and quality management studies in LTC settings. Individual-level variables (sex, age, economic eligibility, LTC approval grade, Charlson Comorbidity Index [CCI], and length of stay) have been validated as predictors of UTI and pressure injury experience and were extracted directly from resident demographic and clinical records.
Age was included as a continuous variable. Sex was categorized as a dichotomous variable indicating whether the resident was female. Economic eligibility for LTC insurance was determined based on an individual’s income and asset levels. Economic eligibility was coded into four categories: National Basic Livelihood Security beneficiaries (lowest economic status) as 0, Medical Aid beneficiaries as 1, individuals eligible for reduced legal LTC insurance out-of-pocket payments as 2, and general beneficiaries without additional public support (highest economic status) as 3.
The LTC approval grades are classified from Grades 1 to 5 by the LTC grade approval board’s decision based on standardized LTC needs assessment scores and doctors’ opinions, with Levels 1 and 5 indicating the highest and lowest LTC needs, respectively. For the analytic model used in this study, LTC approval grades were reverse-coded to reflect care needs such that Level 1 (highest care needs) was coded as 5 and Level 5 (lowest care needs) was coded as 1. The length of nursing home stay was defined as the total number of months residents resided in a nursing home during the 2-year study period.
Baseline health status was measured using CCI, calculated from national health insurance claim records during the year preceding nursing home admission. We calculated CCI using national health insurance claims data from the 12-month period preceding the LTCF admission date, based on information in the NHIS claims database for LTCF residents whose admission date fell between January 1, 2016, and December 31, 2017. CCI was calculated according to the method proposed by Quan et al. [24,25]. For each subject, CCI was derived using primary diagnosis codes from inpatient claims occurring within 1 year prior to the index admission date of the last continuous residency episode as of December 31, 2017. CCI assigns weighted scores to 19 comorbidity categories—ranging from 1 point (e.g., chronic pulmonary disease, rheumatologic disease) to 2 points (e.g., congestive heart failure, dementia, peripheral vascular disease), 4 points (e.g., moderate or severe liver disease, acquired immune deficiency syndrome/human immunodeficiency virus), and 6 points (metastatic solid tumor)—and sums these weights to adjust for mortality risk. The KCD-7 code definitions for each category are presented in the Appendix 1.
Institutional-level variables (ownership type, facility size, registered nurse [RN] staffing, contracted physician visits, additional staffing, and LTCF evaluation indicators) demonstrated significance in institutional theory–based research and were operationalized using 2018 LTCF evaluation data, staffing records from LTC insurance claim records, and LTCF registry. These facility attributes capture organizational capacity for quality management, infection control, and Pressure Injury Prevention, facilitating analysis of their independent associations with adverse event experiences.
Institutional-level independent variables included general facility characteristics, healthcare services, staffing attributes, and LTCF evaluation ratings. General facility characteristics included ownership type, location, size, and the proportion of residents classified as LTC Grades 1 and 2. Ownership type was categorized into three types: other/individual ownership was coded as 1, corporate coded 2, and public ownership was coded as 3. The ‘other’ category in ownership type encompasses facilities that cannot be classified as individual, corporate, or public and includes only nine facilities (0.21%) in the study sample. The ‘other’ category was combined with individual-owned facilities for analysis since they constituted the largest group. Facility locations were classified as rural areas (county, coded as 1), mid-sized cities (city, coded as 2), and metropolitan cities (district, coded as 3). Facility size was defined as the number of licensed beds: fewer than 30 beds (coded as 1), 30~99 beds (coded as 2), and 100 or more beds (coded as 3). The proportion of residents classified as LTC Grades 1 and 2 was calculated as the percentage of residents among the total number of residents at the end of each month of the study period.
Healthcare services and staffing attributes included the presence of RNs, number of monthly visits by contracted physicians, number of affiliated medical institutions, additional staffing beyond minimum legal requirements, and night-shift staffing. The presence of RNs was assessed as a dichotomous variable indicating whether the facility employed at least one RN (coded 1) or not (coded 0). Monthly visits by contracted physicians were defined as the number of visits to a nursing home by contracted physicians per month. The number of affiliated medical institutions was defined as the total number of medical institutions with formal agreements for resident care. Additional staffing was determined by whether the facilities employed direct-care staff (personal care workers, RN, nurse assistants, physical therapists, occupational therapists, and social workers) beyond the legally mandated minimum staffing levels.
The LTCF evaluation ratings were derived from the 2018 LTCF evaluation, ranging from A (highest quality) to E (lowest quality), and coded numerically from 5 (rating A) to 1 (rating E). The 2018 LTCF evaluation was a statutory triennial assessment conducted under the LTC Insurance Act, designed to enhance the quality of care and safeguard beneficiary choice. Evaluation teams of two assessors visited each facility on-site—accompanied by its administrator or representative—and conducted assessments based on standardized indicators and a detailed evaluation. Five major domains were assessed: institutional management (governance, human resource management, information management, quality management), environment and safety (infection control, facility and equipment management, safety management), rights and responsibilities (beneficiary rights, facility responsibilities), care delivery process (service initiation, care planning, care delivery), and care delivery outcomes (satisfaction assessment, beneficiary status). Domain scores were weighted and summed to yield a total score out of 100. In 2018, 48 indicators were applied, and facilities were assigned grades from A (excellent) to E (poor) based on their total scores. Results were published on the LTCI website.
UTI care processes in nursing homes were evaluated using two indicators from the LTCF evaluation ratings, including Excretion Management and Indwelling Catheter Status. LTCF evaluation indicator, Excretion Management, assessed whether excretion services were provided according to beneficiaries’ functional status using three criteria: (1) identification of beneficiaries requiring Excretion Management through daily status monitoring; (2) timely and appropriate interventions (immediate diaper changes, provision of mobile/portable commodes); (3) management of indwelling catheters per physician order. The scores ranged as follows: Excellent (3 points) for meeting all criteria, Good (2.25 points) for meeting criteria (1)+(2) or (1)+(3), Moderate (1.5 points) for meeting criterion (1) only, and Failing (1 point) for meeting no criteria.
The other LTCF evaluation indicator, Indwelling Catheter Status, assessed whether catheter insertion rates were low and removal rates were high using two criteria: (1) post-admission catheter insertion rate below 3% and (2) catheter removal rate due to improved voiding function of at least 20%. Scoring included Excellent (2 points) for meeting both criteria, Good (1.5 points) for meeting partial criteria, and Failing (0 points) for meeting no criteria.
The pressure injury care process in nursing homes was evaluated using the LTCF evaluation indicator of Pressure Injury Prevention, which measures whether appropriate management is provided for beneficiaries at risk of pressure injury development. This indicator reflects international guidelines emphasizing comprehensive risk assessment, preventive interventions, such as regular repositioning, and daily monitoring of high-risk individuals. The indicator comprises three criteria: (1) quarterly assessment of pressure injury risk in at-risk beneficiaries; (2) implementation of preventive efforts (including repositioning at least every 2 hours, even during sleep, and provision of pressure-relieving devices); (3) daily observation and documentation of pressure injury occurrence in high-risk beneficiaries. The scoring system was as follows: Excellent (2 points) if all criteria were met, Good (1.5 points) if two criteria were met, Moderate (1 point) if one criterion was met, and Failing (0 points) if none were met.
4. Data Analysis
Multilevel logistic regression analyses were conducted using STATA version 15, considering the nested structures of residents within the facilities. The Intraclass Correlation Coefficient (ICC) was calculated using an unconditional model to determine the proportion of variance attributable to institutional-level factors. Subsequently, conditional models incorporating all independent variables were estimated for each adverse event.
Variance Inflation Factor (VIF) tests for model 2 which include both individual- and institutional-variables were conducted to assess multicollinearity among variables. In the analysis of UTI care process indicators, the mean VIF was 3.93, with VIFs of 1.33 and 1.02 for LTCF evaluation indicator of Excretion Management and Indwelling Catheter Status, respectively. In the analysis of pressure injury care process indicators, the mean VIF was 4.04, and the VIF for LTCF evaluation indicator of Pressure Injury Prevention was 1.25. All VIF values were below 10, indicating no significant multicollinearity issues.
RESULTS
1. General Characteristics
As shown in Table 1, 122,312 LTCF residents were included in the analysis. The sex distribution showed a predominance of females (n=97,312, 79.6%) compared to males (n=25,000, 20.4%). The age distribution was heavily concentrated in the older age groups, with 103,229 residents (84.4%) aged 75 years or older, specifically 51,986 (42.5%) in the 75~84 years age group, and 51,243 (41.9%) aged 85 years or older. Economic eligibility status showed that general beneficiaries were the largest group (n=64,566; 52.8%), followed by reduced-rate beneficiaries (n=30,345; 24.8%). Baseline health status, measured using the CCI, indicated that most residents (n=104,265; 85.2%) had no significant comorbidities (CCI=0), whereas 15,137 residents (12.4%) had a CCI score of 2. The length of stay showed that nearly half of the residents (n=60,140, 49.2%) had been in the facility for 2 years or longer.
Table 2 provides descriptive statistics of the institutional characteristics of 4,242 LTCFs included for this study. Facility ownership was distributed as follows: individual or other ownership accounted for 67.8% (n=2,876), corporate ownership for 29.6% (n=1,255), and public ownership for 2.6% (n=111). Geographically, mid-sized cities were the most common (56.4%), followed by metropolitan cities (26.1%) and rural areas (17.5%). Licensed bed capacity was disproportionately distributed: <30 beds (65.6%), 30~99 beds (29.5%) and ≥100 beds (4.9%). RNs were employed in 19.6% of the facilities and 63.2% had additional direct-care workers. Residents with LTC Grades 1~2 averaged 34.34%±18.24%, and contracted physician visits per month averaged 2.26±2.39.
The LTCF evaluation ratings were classified as Grade C (24.5%), Grade B (21.6%), Grade E (20.7%), Grade D (19.7%), and Grade A (13.5%). The care process indicators related to the corresponding UTIs and pressure injury performance varied markedly. For Excretion Management indicator, 51.8% of the facilities received failing ratings, whereas 41.2% achieved excellent ratings. Regarding the Indwelling Catheter Status indicator, 90.6% of the facilities achieved excellent performance. Regarding Pressure Injury Prevention indicator, 80.4% of the facilities demonstrated good or excellent performance (49.9%, excellent; 30.5%, good).
Based on NHIS data records during the study period, 38,030 residents (31.1%) had UTI experience, and 8,177 residents (6.7%) had pressure injury experience, as shown in Table 3. The experience of UTI was approximately 4.6 times higher than that of pressure injury.
2. Unconditional Model
Variance decomposition for UTI and pressure injury experience was analyzed using unconditional models for multilevel analysis (Table 4). Random effects analysis revealed institutional-level variance of 0.40 (95% confidence interval [95% CI]=0.37~0.43, p<.001) for UTI experience and 0.41 (95% CI=0.36~0.46, p<.001) for pressure injury experience. The ICC was 11% for UTI experience and 11% for pressure injury experience, indicating the proportion of total variance explained at the institutional level. The Likelihood ratio test result indicates significant between-facility clustering, demonstrating the necessity of multilevel modeling with facility-level random effects.
3. Multivariate Model
Table 5 shows the results of the multivariate analysis of individual and institutional factors related to the experience of UTIs. In both Model 1 and Model 2, individual-level variables—sex, economic eligibility, LTC approval grade, CCI, and months of stay—were significantly associated with UTI experience. Female residents exhibited 16% higher UTI experience compared with males (Model 2: odds ratio [OR]=1.16, 95% CI=1.13–1.21, p<.001). Residents eligible for reduced out-of-pocket payments had an 18% lower UTI experience compared with National Basic Livelihood Security beneficiaries (OR=0.82, 95% CI=0.79~0.86, p<.001), and general beneficiaries showed an 8% reduction (OR=0.92, 95% CI=0.89~0.95, p<.001). Lower LTC approval grades were associated with stepwise increases in UTI experience, with Grade 1 residents having nearly three-fold higher experience than Grade 5 (OR=2.97, 95% CI=2.56~3.44, p<.001). Each 1-point increase in CCI corresponded to a 9% increase in UTI experience (OR=1.09, 95% CI=1.07~1.11, p<.001), and longer stay duration was similarly associated with higher experience rates (OR=1.03, 95% CI=1.03~1.03, p<.001).
Among institutional-level variables, facilities with 30~99 beds (OR=1.12, 95% CI=1.06~1.19, p<.001) and ≥100 beds (OR=1.16, 95% CI=1.03~1.29, p=.01) experienced higher UTI rates. A greater proportion of residents at LTC Grades 1~2 was associated with 36% higher UTI experience (OR=1.36, 95% CI=1.17~1.59, p<.001). Increased contracted physician visits (OR=1.02, 95% CI=1.01~1.03, p<.001) and additional staffing beyond minimum requirements (OR=1.11, 95% CI=1.05~1.18, p<.001) were also linked to higher UTI experience. Conversely, the LTCF evaluation indicator of Indwelling Catheter Status reduced UTI experience by 15% (OR=0.85, 95% CI=0.77~0.95, p<.001).
Likelihood ratio tests indicated both models were statistically significant (Model 1: χ2=4,251.51, p<.001; Model 2: χ2=4,008.38, p<.001), with Model 2 showing a slightly improved fit (Akaike Information Criterion [AIC]=143,238.1).
Table 6 presents the multivariable analysis individual- and institutional-level determinants associated with pressure injury experience. In both Model 1 and Model 2, individual-level variables—sex, economic eligibility, LTC approval grade, and CCI—were significantly associated with pressure injury experience. Female residents exhibited 27% lower experience compared with males in Model 2 (OR=0.73, 95% CI=0.69~0.78, p<.001). Residents eligible for reduced out-of-pocket payments showed a 29% reduction in pressure injury experience relative to National Basic Livelihood Security beneficiaries (OR=0.71, 95% CI=0.66~0.76, p<.001), and general beneficiaries experienced an 11% reduction (OR=0.89, 95% CI=0.84~0.95, p<.001). Higher LTC approval grades were associated with stepwise increases in experience: Grade 1 residents experienced pressure injuries 8.34 times more often than Grade 5 residents (95% CI=5.96~11.68, p<.001). Each 1-point increase in CCI corresponded to a 9% increase in experience (OR=1.09, 95% CI=1.05~1.12, p<.001).
At the institutional level, corporate or public ownership was protective (OR=0.85, 95% CI=0.78~0.92, p<.001; OR=0.80, 95% CI=0.67~0.97, p=.02), and a higher proportion of residents at LTC Grades 1~2 was associated with 21% lower experience (OR=0.79, 95% CI=0.64~0.98, p=.03). Facilities rated B and A showed 14% (OR=0.86, 95% CI=0.77~0.97, p=.02) and 16% (OR=0.84, 95% CI=0.74~0.97, p=.01) lower experience, respectively. Notably, the LTCF evaluation indicator of Pressure Injury Prevention reduced pressure injury experience by 12% (OR=0.88, 95% CI=0.82~0.95, p<.001).
Likelihood ratio tests indicated both models were highly significant (Model 1: χ2=4,600.73, p<.001; Model 2: χ2=1,019.67, p<.001). Model 2 demonstrated slightly improved fit (AIC=55,850.46).
DISCUSSION
This study comprehensively analyzed individual- and institutional-level factors influencing UTI and pressure injury experiences in Korean LTCFs. It examined the comprehensive risk variables available in administrative datasets and determined the organizational factors that could be directly targeted through prevention-oriented policy measures using a multilevel approach.
Based on a multilevel analysis of 122,312 residents, the UTI experience rate (31.09%) was approximately 4.6 times higher than the pressure injury experience rate (6.69%). Individual factors related to UTI and pressure injury experiences among Korean nursing home residents varied considerably even after adjusting for institutional factors.
The opposite effects of sex on these two outcomes are noteworthy. The results showing a higher UTI risk and lower pressure injury risk in females are consistent with those of the previous literature [19,20]. The anatomical characteristics of females, particularly a shorter urethra, contribute to increased UTI vulnerability, while males may face a higher tissue damage risk at pressure points due to greater body weight and muscle mass, supporting previous research findings. Age emerging as a significant risk factor only for pressure injuries can be attributed to the age-related decline in skin elasticity and tissue regeneration capacity [26].
Higher LTC needs, presented as lower LTC grades, showed stepwise risk increases for both outcomes, with more pronounced patterns for pressure injuries. This reflects a direct association between functional decline and reduced mobility, which has a greater impact on the development of pressure injuries. The protective effects of better economic eligibility on both outcomes likely reflect better nutritional status and healthcare accessibility.
Our findings indicate that the experiences of UTIs and pressure injuries among residents in Korean LTCFs are associated with various institutional factors. The analysis using the unconditional empty model revealed that the ICC was 11% for both UTIs and pressure injuries. This suggests that approximately 11% of the variance in these adverse outcomes can be attributed to differences between institutions. These results highlight the potential for managing and reducing the risk of UTIs and pressure injuries by addressing specific institutional characteristics. This underscores the importance of organizational-level interventions and policies aimed at improving care quality and resident outcomes in LTC settings.
The finding that facility size was a risk factor only for UTIs is notable. This may be attributed to the increased risk of cross-infection in larger facilities and difficulties in providing individualized care. Conversely, the contradictory results, where higher proportions of high-acuity residents had increased UTI risk but reduced pressure injury risk, suggest that intensive Pressure Injury Prevention management is being implemented in high-acuity patients.
The counterintuitive association between increased physician visits, additional staffing, and higher UTI risk can be explained through two complementary theoretical frameworks: structural healthcare system constraints and methodological confounding factors. Structural healthcare system limitations provide the primary explanation for these unexpected findings. The Korean LTC regulatory framework fundamentally restricts contracted physicians to monitoring and referral functions rather than direct therapeutic interventions. Medical interventions in Korean LTCFs are strictly regulated, limiting contracted physicians to identifying health problems and facilitating transfers to acute care settings or connecting residents with National Health Insurance home nursing services [27]. This regulatory constraint is reinforced by institutional philosophy, where facility administrators conceptualize LTCFs as residential rather than medical settings, prioritizing hospital transfers over in-facility medical management when health conditions deteriorate [27]. This potentially explains why increased allocation of medical resources fails to improve infection outcomes.
Methodological confounding mechanisms offer additional explanation. Surveillance bias represents a critical factor wherein facilities with enhanced medical resources demonstrate superior UTI detection capabilities through more rigorous monitoring protocols and diagnostic practices, resulting in artificially elevated reporting rates that reflect diagnostic sensitivity rather than actual infection incidence. Reverse causality presents an equally compelling explanation, suggesting that medical resource allocation functions as a reactive strategy where facilities caring for higher-acuity residents receive additional physician visits and staffing in response to existing patient complexity and infection risk, rather than as a preventive measure.
While international evidence demonstrates that increased RN staffing in nursing homes improves resident health outcomes, including reduced UTIs, pressure injuries, weight loss, and enhanced activities of daily living, our study failed to identify significant correlations between nursing staff levels and health outcomes in Korean facilities. This finding is likely attributable to critically insufficient staffing levels, as 2019 OECD data indicate that Korea maintains only 0.1 nurses per 100 elderly residents in LTC institutions compared to Switzerland (3.1), Germany (1.6), and the United States (0.9), with the 2022 LTC survey revealing that RNs comprise merely 0.3 positions among an average of 26.3 total staff in Korean nursing homes [28].
The critical context of nursing staffing in Korean LTCFs must be considered when interpreting these results. The fact that only 0.3 RNs are employed per 26.3 LTC staff members indicates that a majority of Korean nursing homes operate without any professional nursing staff, instead relying on nursing aids or personal care workers to perform nursing functions. Consequently, the shortage of specialized personnel impedes professional infection management, hindering proper urinary management and personal hygiene guidance, delaying early recognition and response to UTI symptoms, increasing infection risk through non‐expert catheter care, and undermining the application of systematic infection‐control protocols.
Rather than merely increasing overall staffing levels, mandatory deployment and augmentation of RNs with specialized infection‐control training are essential; ensuring dedicated expertise within facilities can facilitate structured urinary care programs, prompt symptom detection and rapid referral, and ultimately achieve substantial reductions in UTI incidence. Thus successful implementation of the Korean government’s planned mandatory nurse deployment policy for facilities with 50 or more residents is essential to achieve internationally comparable care quality standards.
The analysis of institutional-level variables reveals important insights into targeted quality management approaches. The results showed that the Indwelling Catheter Status indicator reduced UTI risk by 14%, and the Pressure Injury Prevention indicator reduced pressure injury risk by 12%, emphasizing the importance of condition-specific tailored quality management. Condition-specific LTCF evaluation indicators demonstrated protective effects exclusively for their corresponding outcomes—the Indwelling Catheter Status indicator for UTIs and the Pressure Injury Prevention indicator for pressure injuries—validating the specificity of tailored quality management strategies.
Our findings suggest that improvements in facility-level service quality outcomes may be fundamentally linked to national-level policies designed to promote best practices. International examples demonstrate various systematic approaches to quality improvement in LTC settings. Germany has developed expert care standards for key clinical conditions including pressure injuries, UTIs, and other adverse events, mandating their implementation in nursing facilities and incorporating them into service quality evaluations [29]. Similarly, the US Centers for Medicare & Medicaid Services nursing home evaluation system monitors incidence rates of major clinical outcome indicators and integrates these metrics into certification processes and performance-based payment systems [30].
Korea has implemented a similar approach through the LTCF evaluation program, jointly administered by the Ministry of Health and Welfare and the NHIS to promote quality improvement in nursing facilities systematically. All Korean LTCFs are mandated to undergo evaluation, with results publicly disclosed to ensure transparency and accountability. The system provides financial incentives based on evaluation outcomes, motivating facilities to achieve higher quality scores [30]. Given these policy frameworks and our study findings, government agencies, insurers, and academic institutions should collaborate with practitioners to develop evidence-based care guidelines and establish comprehensive education and training programs for facility staff. This collaborative investment is crucial for enabling LTCFs to implement best practices and effectively prevent adverse events, such as UTIs and pressure injuries.
This study has limitations, as it was a retrospective study using health insurance claims data, restricting causal inference. Owing to the nature of claims data, information on clinical severity or specific care processes is limited, and asymptomatic infections or mild pressure injuries may have been underestimated.
This study analyzed the patterns of UTI and pressure injury experiences in a large, representative sample of Korean nursing home residents by linking LTCI and NHI claims data. Operational definitions based on public administrative data have enabled comprehensive population-level analysis of adverse events. This approach addresses the limitations of previous Korean studies that relied on small facility-specific samples. Linked administrative datasets provide a more robust foundation for examining the epidemiology and risk factors of these conditions in LTC settings.
CONCLUSION
This study comprehensively analyzed individual- and institutional-level factors associated with UTIs and pressure injuries among 122,312 Korean nursing home residents, revealing UTI experience rates of 31.09% and pressure injury experience rates of 6.69%. Individual-level factors including sex, economic eligibility, LTC approval grade, and CCI significantly influenced both outcomes, with females showing higher UTI risk but lower pressure injury risk. At the institutional level, facility size, resident acuity, and condition-specific quality management indicators emerged as significant determinants, with the ICC of 11% for both outcomes indicating substantial facility-level variance. The finding that condition-specific quality indicators showed protective effects exclusively for their corresponding outcomes validates the effectiveness of targeted quality management approaches. These results demonstrate that both individual characteristics and organizational factors contribute significantly to adverse event experience, supporting the need for multilevel intervention strategies.
Despite study limitations including reliance on administrative data and potential surveillance bias, this population-based analysis provides valuable insights for developing evidence-based prevention strategies. Policymakers should make an effort to prioritize implementing mandatory RN deployment policies and strengthening physician support systems to achieve internationally comparable care quality standards. Government agencies, insurers, and academic institutions must collaborate to develop evidence-based care guidelines and comprehensive education programs for facility staff. Future research should incorporate prospective studies examining qualitative aspects of care processes and randomized controlled trials evaluating customized prevention interventions. A systematic extension of rigorous institutional evaluations incorporating condition-specific quality indicators is essential for substantial improvements in LTC service quality and safety.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation 
Print 
