Scholarlink

Background <p>Fungal infection remains a recurring concern during inpatient care for acute myeloid leukemia (AML), but many hospitals lack local estimates of the bed-day and cost burden associated with fungal-infection documentation. We estimated excess length of stay (LOS) and hospitalization cost associated with routine-care documented probable or suspected fungal infection (PFI) among AML inpatient encounters.</p> Methods <p>We conducted a retrospective encounter-level study using routinely collected 2020–2025 AML inpatient data from the electronic medical record database of Ganzhou People’s Hospital. PFI was identified from prespecified fungal-infection terms in diagnosis text and microbiology fields (excluding records with an explicitly negative fungal result) to define routine-care documented PFI. Outcomes were LOS and total hospitalization cost in Chinese yuan (Renminbi; RMB, also CNY). The primary estimand was the average treatment effect among the treated (ATT). We fitted crude, propensity score matching, inverse probability of treatment weighting (IPTW), and doubly robust models. Sensitivity analyses evaluated alternative weighting and matching choices, robust uncertainty, PFI documentation source, patient-level clustering, cost missingness, and the feasibility of an early-PFI definition.</p> Results <p>The analytic cohort comprised 2,686 AML inpatient encounters; 661 (24.6%) were classified as PFI and 2,025 as non-PFI. Median LOS was 18.97 days in the PFI group and 7.06 days in the non-PFI group. Among encounters with observed cost, median hospitalization cost was 24,888.24 RMB in the PFI group and 7,681.44 RMB in the non-PFI group. In the primary IPTW-ATT analysis, PFI was associated with 10.17 excess hospital days (95% CI, 8.49 to 11.85) and 20,542.01 RMB excess hospitalization cost (95% CI, 16,361.90 to 24,722.11). Doubly robust ATT estimates were similar: 10.21 days for LOS and 20,613.70 RMB for cost. Heteroskedasticity-robust (non-clustered) standard errors yielded directionally consistent intervals for IPTW-ATT: LOS 10.17 days (95% CI, 8.81 to 11.53) and cost 20,542 RMB (95% CI, 17,205 to 23,879). Cost was missing in 506 encounters (18.8%); cost-missing encounters differed modestly from cost-observed encounters.</p> Conclusions <p>In this single-hospital AML inpatient dataset, documented PFI was associated with a higher-burden hospital course, including longer LOS and higher observed cost.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Hospital length of stay and cost associated with routine-care documentation of probable or suspected fungal infection among patients hospitalized with acute myeloid leukemia: a propensity score analysis

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年7月22日

查看全文
  • Chengming He,
  • Xufeng Liu,
  • Biqing Wang,
  • Le Hu,
  • Jingdong Zhang,
  • Xianying Zeng,
  • Bolin Wan

摘要

Background

Fungal infection remains a recurring concern during inpatient care for acute myeloid leukemia (AML), but many hospitals lack local estimates of the bed-day and cost burden associated with fungal-infection documentation. We estimated excess length of stay (LOS) and hospitalization cost associated with routine-care documented probable or suspected fungal infection (PFI) among AML inpatient encounters.

Methods

We conducted a retrospective encounter-level study using routinely collected 2020–2025 AML inpatient data from the electronic medical record database of Ganzhou People’s Hospital. PFI was identified from prespecified fungal-infection terms in diagnosis text and microbiology fields (excluding records with an explicitly negative fungal result) to define routine-care documented PFI. Outcomes were LOS and total hospitalization cost in Chinese yuan (Renminbi; RMB, also CNY). The primary estimand was the average treatment effect among the treated (ATT). We fitted crude, propensity score matching, inverse probability of treatment weighting (IPTW), and doubly robust models. Sensitivity analyses evaluated alternative weighting and matching choices, robust uncertainty, PFI documentation source, patient-level clustering, cost missingness, and the feasibility of an early-PFI definition.

Results

The analytic cohort comprised 2,686 AML inpatient encounters; 661 (24.6%) were classified as PFI and 2,025 as non-PFI. Median LOS was 18.97 days in the PFI group and 7.06 days in the non-PFI group. Among encounters with observed cost, median hospitalization cost was 24,888.24 RMB in the PFI group and 7,681.44 RMB in the non-PFI group. In the primary IPTW-ATT analysis, PFI was associated with 10.17 excess hospital days (95% CI, 8.49 to 11.85) and 20,542.01 RMB excess hospitalization cost (95% CI, 16,361.90 to 24,722.11). Doubly robust ATT estimates were similar: 10.21 days for LOS and 20,613.70 RMB for cost. Heteroskedasticity-robust (non-clustered) standard errors yielded directionally consistent intervals for IPTW-ATT: LOS 10.17 days (95% CI, 8.81 to 11.53) and cost 20,542 RMB (95% CI, 17,205 to 23,879). Cost was missing in 506 encounters (18.8%); cost-missing encounters differed modestly from cost-observed encounters.

Conclusions

In this single-hospital AML inpatient dataset, documented PFI was associated with a higher-burden hospital course, including longer LOS and higher observed cost.