CoChat
Sign Up
Cited by 0Healthcare Quality and ManagementDialysis and Renal Disease ManagementNursing education and management

Longitudinal Analysis and Predictive Modeling of Nursing‐Sensitive Quality Indicators in Hemodialysis

Open access
Tang, Sikai, Li, Qiao, Liu, Li, He, Li, Zhang, Yingjun, Chen, Lin·2026·Journal of Nursing Management
DOI 10.1155/jonm/8833031PMC PMC13415755
Open this paper in CoChat — free account

Abstract

Objective: Nursing‐sensitive quality indicators (NSQIs) are essential for evaluating and improving the quality of hemodialysis (HD) care, yet long‐term monitoring of their trends and interrelationships remains limited. This study aimed to examine longitudinal trends and interrelationships among NSQIs in an HD unit and to apply analytical methods to support nursing quality management. Methods: A 4‐year longitudinal retrospective study was conducted in a tertiary hospital HD unit. Ten NSQIs (two structural indicators, three process indicators, and five outcome indicators) were monitored monthly from January 2022 to December 2025. Data were analyzed using descriptive statistics, correlation analysis, principal component analysis (PCA), and autoregressive integrated moving average (ARIMA) modeling. Results: Structural and process indicators remained generally stable or improved over the 4 years. Among outcome indicators, incidence of hypotension in HD (O1) and patient satisfaction (O5) were stable; incidence of coagulation during extracorporeal circulation (O2) increased initially and then declined; dialysis period weight control success rate (O3) decreased and then partially recovered; incidence of central venous catheter infection (O4) decreased markedly. Spearman correlation showed a positive association between the ratio of blood purification specialist nurses to general nurses (S2) and O3, and an inverse relationship between process indicators and O4. PCA explained 49.5% of the total variance and revealed an annual evolution in the structure of nursing quality indicators. The final ARIMA models for O1–O4 all passed the Ljung–Box test ( p > 0.05). Six‐month forecasts indicated that O1 and O3 would remain nearly constant, O2 would increase initially and then stabilize, and O4 would rise slowly. Conclusion: Continuous monitoring of NSQIs helps identify actionable areas for quality improvement in HD care. The structure–process–outcome framework effectively captured dynamic trends in nursing quality, revealing distinct patterns across indicator types and highlighting areas requiring targeted intervention. Predictive modeling supports proactive management and adjustable interventions. Implications for Nurse Leaders: Nurse leaders can use longitudinal NSQI data to guide staffing decisions, prioritize process‐centered quality improvement, and design targeted interventions based on outcome trends. Time‐series models help move from reactive to predictive management, making it easier to anticipate trend shifts, ultimately improving patient outcomes in HD care.

AI snapshot

Written from the paper’s full text.

TL;DR

This 4-year single-center retrospective study monitored 10 nursing-sensitive quality indicators in a hemodialysis unit and found that structural/process indicators were generally stable or improved while outcome trends varied. PCA and ARIMA modeling showed shifting indicator structure over time and produced exploratory 6-month forecasts, with coagulation risk and catheter infection requiring targeted attention.

Key findings

  • Across 48 months, mean S1 was 4.69 ± 0.15, median S2 was 75.41% (IQR 4.99), and process indicators were high: P1 97.77 ± 0.90%, P2 97.56 ± 0.88%, and P3 97.05 ± 0.60%.
  • Outcome indicators showed distinct patterns: O1 was 1.670 ± 0.53%, O2 was 0.313% (IQR 0.17), O3 was 84.35 ± 5.20%, O4 was 0.018‰ (IQR 0.03), and O5 was 99.80 ± 0.12%; O4 fell from 0.047‰ in 2022 to 0.000‰ in 2025, while O3 decreased to 80.53% in 2024 and recovered to 86.61% in 2025.
  • Spearman correlations showed S2 positively associated with O3 (r = 0.68, p < 0.001) and inversely associated with O1 (r = −0.49, p < 0.001) and O2 (r = −0.32, p < 0.05); P1 and P2 were inversely correlated with O4 (r = −0.47 and −0.52, respectively, p < 0.001).
  • Pooled PCA explained 49.5% of total variance, with yearly PCAs explaining 51.4%–57%; final ARIMA models for O1–O4 all passed the Ljung–Box test (p > 0.05), and 6-month forecasts predicted O1 constant at 2.28%, O3 constant at 87.27%, O2 rising then stabilizing around 0.316%, and O4 rising slowly from 0.006‰ to 0.019‰.

Methods

  • Retrospective longitudinal study in the hemodialysis unit of West China Hospital of Sichuan University, using monthly data from January 2022 to December 2025; as of December 2025 the unit had 66 dialysis machines, 61 registered nurses (46 certified blood purification specialist nurses), 315 outpatients, and 5,447 dialysis sessions.
  • Ten NSQIs were monitored: structural S1 (nurse–patient ratio) and S2 (specialist-to-general nurse ratio); process P1 (catheter/puncture needle standard fixation), P2 (hand hygiene execution), and P3 (AVF rope-ladder puncture); outcome O1 (hypotension), O2 (extracorporeal circulation coagulation), O3 (dialysis period weight control success), O4 (central venous catheter infection), and O5 (patient satisfaction).
  • Analysis included descriptive statistics, Shapiro–Wilk tests, repeated measures ANOVA or Friedman tests with post hoc comparisons, Spearman correlation, PCA, and Box–Jenkins ARIMA modeling with ADF tests, AIC-based model selection, Ljung–Box residual diagnostics, and 6-month forecasts.

Limitations

  • Stated in the paper: the single-center retrospective design limits generalizability; the NSQIs were developed and refined within one center and require external validation; observational data preclude causal inference and unmeasured confounders may remain.
  • Stated in the paper: rare events such as falls and unplanned extubation were excluded; the relatively short time series limits long-term forecasting robustness, so the 6-month forecasts are exploratory; multiple concurrent quality improvement interventions were not evaluated under controlled conditions, making individual effects difficult to isolate.
  • Evident from the text: seasonal ARIMA models were not used because of the small sample size, and O4 contained zero values, so MAPE was not reported for that indicator.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1155/jonm/8833031

Cites

22 of 22 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Claims the paper makes

5 of 12 cited sources checked · 3 supported · 0 flagged · accuracy not yet measured

  • Premisecitation

    The highest relative IDWG category (≥ 5.7%) is associated with significantly increased mortality risk.

    “the highest relative IDWG category (≥ 5.7%) has been associated with significantly increased mortality risk [26]”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites [26] Wong M. M., McCullough K. P., Bieber B. A. et al., Interdialytic Weight Gain: Trends, Predictors, and Associated Outcomes in the International Dialysis Outcomes and Practice Patterns Study (DOPPS), American Journal of Kidney Diseases. (2017) 69, no. 3, 367– 379, 10.1053/j.ajkd.2016.08.030. 27866963 source
  • Findingcitation

    Our S1 is lower than those commonly observed in South Korea (1:5-1:6).

    “Compared internationally, our S1 is higher than Australian standards (1:3‐1:4) [20] but lower than those commonly observed in South Korea (1:5‐1:6) [21].”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites [21] Lee Y. K., Kim K., and Kim D. J., Current Status and Standards for Establishment of Hemodialysis Units in Korea, The Korean Journal of Internal Medicine. (2013) 28, no. 3, 274– 284, 10.3904/kjim.2013.28.3.274. 23682218 PMC3654122 source

      The Proposed Facility Standards for Dialysis Units section recommends no more than five daily dialysis sessions per nurse, with an allowance of six; the Dialysis Unit Accreditation Project section reports an average of 4.97 sessions per nurse in one project (page numbers unavailable). These figures are daily workload measures, not evidence that a 1:5–1:6 nurse-to-patient ratio was commonly observed across South Korean units. (LLM, conf=0.86)

      Experimental — accuracy not yet measured
  • Findingcitation

    Our S1 is higher than Australian standards (1:3-1:4).

    “Compared internationally, our S1 is higher than Australian standards (1:3‐1:4) [20] but lower than those commonly observed in South Korea (1:5‐1:6) [21].”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites [20] Damasiewicz M. J. and Polkinghorne K. R., Global Dialysis Perspective: Australia, Kidney360. (2020) 1, no. 1, 48– 51, 10.34067/kid.0000112019. 35372853 PMC8808481 source
  • Findingown data

    The pooled principal component analysis explained 49.5% of the total variance.

    “PCA explained 49.5% of the total variance”
    • Also stated in: § 3. Results · plain
      “which together explained 49.5% of the total variance.”
    Quote in the paperNot a substantial claim
  • Findingown data

    Incidence of central venous catheter infection (O4) decreased markedly over the 4 years.

    “incidence of central venous catheter infection (O4) decreased markedly.”
    Quote in the paperNot a substantial claim
  • Classified as probably not a substantial claim. 13
  • May combine two claims. 14

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Previous studies have successfully applied ARIMA models to forecast disease incidence trends, consistently achieving robust fitting and predictive performance.

    “previous studies have successfully applied ARIMA models to forecast disease incidence trends, consistently achieving robust fitting and predictive performance [28]”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites [28] Wang M., Pan J., Li X. et al., ARIMA and ARIMA-ERNN Models for Prediction of Pertussis Incidence in Mainland China From 2004 to 2021, BMC Public Health. (2022) 22, no. 1, 10.1186/s12889-022-13872-9. PMC9338508 35906580 source

      The Introduction (page not available in our copy) reports two previous ARIMA studies of pertussis incidence: one predicted trends with acceptably low errors, and another found its ARIMA model performed best among those tested. That supports successful applications, but not the broader claim of consistently robust fitting and prediction across disease incidence trends. The Results (Model Prediction; page not available) also report that this paper’s ARIMA model performed worse than its combined ARIMA-ERNN model. (LLM, conf=0.84)

      Experimental — accuracy not yet measured

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    Nurses constitute an around-the-clock surveillance system in hospitals, closely monitoring changes in patients' clinical conditions and administering treatments and care as appropriate.

    “Nurses constitute an around‐the‐clock surveillance system in hospitals, closely monitoring changes in patients’ clinical conditions and administering treatments and care as appropriate [22].”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites [22] Lasater K. B., Aiken L. H., Sloane D. et al., Patient Outcomes and Cost Savings Associated With Hospital Safe Nurse Staffing Legislation: An Observational Study, BMJ Open. (2021) 11, no. 12, 10.1136/bmjopen-2021-052899. PMC8655582 34880022 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    For O2, ARIMA(0,1,4) had AIC = −40.80.

    “For comparison, we also examined the models with the next lowest AIC values: ARIMA (0, 1, 4) (AIC = −40.80) and ARIMA (1, 1, 3) (AIC = −40.73).”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    O3 decreased from 2022 to 2024 but partially recovered in 2025.

    “O3 decreased from 2022 to 2024 but partially recovered in 2025, while O4 decreased markedly.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingnone given

    The incidence of O2 increased from 2022 to 2024 but declined in 2025, resulting in a fluctuating trend overall.

    “The incidence of O2 increased from 2022 to 2024 but declined in 2025, resulting in a fluctuating trend overall.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingnone given

    S2 served as a guiding factor among NSQIs in 2022, but its role diminished in 2024 and 2025.

    “S2 served as a guiding factor among NSQIs in 2022, but its role diminished in 2024 and 2025.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    The differenced O3 series had significant partial autocorrelations at lags 1 and 2 in the PACF, then cut off.

    “The ACF of the differenced series showed significant autocorrelations at lags 1, 8, and 10 months, while the PACF showed significant partial autocorrelations at lags 1 and 2 and then cut off.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    For O2, ARIMA(0,1,4) and ARIMA(1,1,3) both had non-significant coefficients and did not yield a better fit than ARIMA(0,1,3).

    “Both had non‐significant coefficients and did not yield a better fit.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    For O2, ARIMA(1,1,3) had AIC = −40.73.

    “For comparison, we also examined the models with the next lowest AIC values: ARIMA (0, 1, 4) (AIC = −40.80) and ARIMA (1, 1, 3) (AIC = −40.73).”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingreasoning

    The rightward shift of samples was accompanied by a marked reduction in O4 but a persistently high O2.

    “the rightward shift of samples was accompanied by a marked reduction in O4 but a persistently high O2”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    For O4, the ACF decayed slowly but became nonsignificant after lag 8.

    “The ACF decayed slowly but became nonsignificant after lag 8, while the PACF cut off after lag 1, suggesting an AR (1) process.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    O3 decreased from 88.41% in 2022 to 80.53% in 2024, but recovered to 86.61% in 2025.

    “O3 decreased from 88.41% in 2022 to 80.53% in 2024, but recovered to 86.61% in 2025”
    Quote in the paperSubstantial claim
    • Rests on an assumption: Interdialytic weight gain (IDWG) directly reflects the volume status of HD patients.

  • May combine two claims.

    Findingown data

    Six-month forecasts for O2 showed a gradual increase from 0.273% in January to 0.328% in February, then stabilized around 0.316% from March to June.

    “Six‐month forecasts showed a gradual increase from 0.273% in January to 0.328% in February, then stabilized around 0.316% from March to June.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingreasoning

    Our study identified that a higher specialist nurse ratio was associated with lower hypotension and lower incidence of coagulation and improved dry weight control.

    “Our study identified that a higher specialist nurse ratio was associated with lower hypotension and lower incidence of coagulation and improved dry weight control.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

    Interpretationnone given

    This consistency reflects successful standardization of catheter fixation, hand hygiene, and rope-ladder puncture techniques.

    “This consistency reflects successful standardization of catheter fixation, hand hygiene, and rope‐ladder puncture techniques.”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: Consistent nursing practices reduce variations in care and support patient safety and satisfaction.

  • Probably not a substantial claim: may not pass S2 (checkable).

    Interpretationreasoning

    Periodic reassessment of staffing policies may be necessary to sustain S2's contribution.

    “implying that periodic reassessment of staffing policies may be necessary to sustain its contribution”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing), S4 (based).

    Interpretationnone given

    These variations likely reflect differences in health care systems, funding models, and patient demographics specific to each country.

    “These variations likely reflect differences in health care systems, funding models, and patient demographics specific to each country.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Interpretationreasoning

    The association between higher S2 values and increased O4 may be attributed to autogenous arteriovenous fistula (AVF) grade management in our unit, in which HD specialist nurses tend to focus more on AVF cannulation, while nonspecialist nurses often place greater emphasis on catheter care.

    “This phenomenon may be attributed to autogenous arteriovenous fistula (AVF) grade management in our unit, in which HD specialist nurses tend to focus more on AVF cannulation. In contrast, nonspecialist nurses often place greater emphasis on catheter care.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Interpretationreasoning

    The single-center retrospective design may limit the generalizability of the results to other dialysis settings with different resource levels or patient populations.

    “the single‐center retrospective design may limit the generalizability of the results to other dialysis settings with different resource levels or patient populations”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductreasoning

    Unmeasured confounders may have influenced the results.

    “unmeasured confounders may have influenced the results”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Interpretationnone given

    The decline in O3 may be partly due to seasonal factors such as increased fluid intake in summer.

    “The decline may be partly due to seasonal factors such as increased fluid intake in summer”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationown data

    ARIMA(1,0,0) was selected as the final model for O4 because it was more parsimonious.

    “The AIC of AR (2) was nearly identical to that of AR (1), so the more parsimonious ARIMA (1, 0, 0) was selected.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Interpretationnone given

    The 6-month forecasts are exploratory.

    “the 6‐month forecasts are exploratory”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationcitation

    Key contributing factors to the downward trend in O4 include national CRBSI prevention policies and our center's specialized nursing management strategy.

    “Key contributing factors include national CRBSI prevention policies [27] and our center’s specialized nursing management strategy.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [27] National Health Commission of the People’s Republic of China, Circular of the General Office of the National Health Commission on Printing and Distributing the National Healthcare Quality and Safety Improvement Goals for 2021: No. 76 of 2021 of the General Office of the NHC [A/OL], 2021, https://www.nhc.gov.cn/yzygj/c100068/202102/116305f4cb174ef59b02be71ac826b6b.shtml.

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Interpretationreasoning

    The NSQIs' applicability and comparability across institutions or cultures require further validation, particularly regarding predictive validity and responsiveness to change.

    “their applicability and comparability across institutions or cultures require further validation, particularly regarding predictive validity and responsiveness to change”
    Quote in the paperNot a substantial claim
25 of 25

  • The paper relies on this without support.

    The observed trends in nursing-sensitive quality indicators reflect changes in nursing quality rather than external factors such as patient case-mix or policy changes.

    “Continuous monitoring of NSQIs helps identify actionable areas for quality improvement in HD care.”

  • The paper relies on this without support.

    The historical patterns in the nursing-sensitive quality indicators will continue over the 6-month forecast period.

    “The forecasts are intended as short‐term trend indicators.”

  • The paper relies on this without support.

    When nurses care for fewer patients at a time, patients are less likely to experience an adverse outcome.

    “When nurses care for fewer patients at a time, they can spend more time at each patient’s bedside, and as a result, patients are less likely to experience an adverse outcome.”

  • The paper relies on this without support.

    The specialist nurse training system remains the primary form of continuing education for nursing professionals in China due to differences in national conditions.

    “Although some APN training initiatives have been introduced in China, the specialist nurse training system, distinct from international models like APN or Clinical Nurse Specialist (CNS), remains the primary form of continuing education for nursing professionals due to differences in national conditions.”

  • The paper relies on this without support.

    Consistent nursing practices reduce variations in care and support patient safety and satisfaction.

    “Consistent nursing practices not only reduce variations in care but also support patient safety and satisfaction.”

  • The paper relies on this without support.

    Conventional coagulation monitoring alone is insufficient for accurately evaluating and managing anticoagulant effects across different agents.

    “conventional coagulation monitoring alone is insufficient for accurately evaluating and managing anticoagulant effects across different agents”

  • The paper relies on this without support.

    Interdialytic weight gain (IDWG) directly reflects the volume status of HD patients.

    “Interdialytic weight gain (IDWG) directly reflects the volume status of HD patients”

  • The paper relies on this without support.

    Inadequate real-time volume monitoring and over-reliance on monthly statistics have led to incomplete quality feedback loops.

    “Inadequate real‐time volume monitoring and over‐reliance on monthly statistics have led to incomplete quality feedback loops.”

  • The paper relies on this without support.

    The ARIMA model is a classical time-series analysis method that offers significant advantages in capturing periodicity and variations within datasets and demonstrates high predictive accuracy.

    “The ARIMA model is a classical time‐series analysis method that offers significant advantages in capturing periodicity and variations within datasets and demonstrates high predictive accuracy.”
9 of 9

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

  • P2-O4 correlation is described with the wrong sign

    Minor

    The text calls the P2-O4 association positive, but reports a negative correlation coefficient (r = −0.52). The abstract also describes the relationship between process indicators and O4 as inverse.

1 of 1

Cites

Flagged first, then by citations

Research works printed but not linked: 5

  1. Chen H., 2023, Application Status and Research Progress of Palliative Care in Patients With End-Stage Renal Disease, Chinese Medical Ethics, 36, 1382
  2. 10.3870/j.issn.1001-4152.2021.14.059 · DOI
  3. 10.12102/j.issn.1009-6493.2024.14.024 · DOI
  4. 10.13912/j.cnki.chqm.2022.29.10.06 · DOI
  5. 10.12102/J.issn.1009-6493.2019.01.026 · DOI

Authors
Tang, S., Li, Q., Liu, L., He, L., Zhang, Y., & Chen, L.
Affiliations
Sichuan University (CN)
Corresponding author
  • Zhang, Yingjun
  • Chen, Lin
ORCID
Venue
Journal of Nursing Management
Publisher
Wiley
Published
2026-01
Volume
2026
Issue
1
Pages
e8833031-e8833031
ISSN
  • 0966-0429 (ISSN-L)
  • 1365-2834
DOI
10.1155/jonm/8833031
PMC
PMC13415755
OpenAlex
W7171660571
Cited by
0
References
22
FWCI
0.00
Type
journal article
Language
en
Open access
Licence
Locations
  • Journal of Nursing Management (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • PubMed Central (submittedVersion, cc-by)
Topics
  • Health Sciences › Health Professions › Health Information Management › Healthcare Quality and Management
  • Health Sciences › Medicine › Nephrology › Dialysis and Renal Disease Management
  • Health Sciences › Nursing › Research and Theory › Nursing education and management
Keywords
nursing-sensitive quality indicators, hemodialysis, nursing quality management, structure-process-outcome, central venous catheter infection, weight management, patient satisfaction, nurse staffing ratios, time series forecasting, quality improvement
MeSH terms
Renal Dialysis, Humans, Longitudinal Studies, Retrospective Studies, Quality Indicators, Health Care

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-10-27.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-27.

Cited sources are checked by gpt-6-sol.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review