Subtype-specific optimal cut-off values of the Berg Balance Scale for predicting independent walking in inpatient stroke rehabilitation: a multicentre cohort study.
Jeong-Soo Kim, In-Geon Hwang, Eu-Jeong Ko
A Berg Balance Scale score at admission predicts independent walking at discharge in stroke rehabilitation.
The useful cut-off is 33 for ischaemic stroke and 12 for haemorrhagic stroke. The single score performs as well as adding age, sex, and other factors to the prediction.
Key findings
1Admission BBS was a significant independent predictor of independent walking at discharge (aOR 1.053, 95% CI 1.030–1.076, p < 0.001), with an AUC of 0.899 (95% CI 0.874–0.921).
2Optimal BBS cut-offs for independent walking differed by stroke subtype: 33 for ischaemic stroke (AUC 0.890, sensitivity 84.1%, specificity 77.0%) and 12 for haemorrhagic stroke (AUC 0.912, sensitivity 96.3%, specificity 75.6%).
3The BBS-only model's discriminative ability was not significantly different from the multivariable model (AUC 0.899 vs 0.931; DeLong p = 0.097), indicating that a single BBS score captures most of the predictive information.
Still to come
How it was doneWhat they foundWhat it means for PTs
Read the rest of this summary
You get three full summaries a month, free, and we do not ask for a card. Search, the TL;DRs and your library stay unlimited either way.
What it does not show
Retrospective design means unmeasured confounders (lesion location, lesion volume, rehabilitation intensity) could not be controlled. Only 565 of 1,117 registry patients (50.6%) were included; the main exclusion reason was missing BBS data (58.5%), which may have disproportionately excluded more severely affected patients. Internal validation only; no external validation in an independent cohort was performed. Single-country cohort (Korea); generalisability to other populations is untested. One centre (Centre A) contributed 56.1% of the sample, and the acute-versus-subacute cut-off difference (10 vs 38) is confounded by centre because the acute group came from a single site. NIHSS data were available for only 21.4% of patients, limiting the severity-adjusted sensitivity analysis. The haemorrhagic stroke cut-off of 12 is based on only 3 of 130 patients below that threshold achieving independent walking, making the estimate unstable.
Declared interests
No conflicts of interest declared. Funded by the National R&D Program for Clinical Field Problem-Solving in Neurorehabilitation Diseases, Ministry of Health and Welfare, Korea (Grant No. RS-2023-00262005). The funding institution had no role in study design, execution, data analysis, interpretation, or dissemination.
The easy way to misread this
Do not treat the haemorrhagic stroke cut-off of 12 as a firm boundary. Only 3 of 130 patients scoring below 12 achieved independent walking, so the odds ratio of 24.778 (95% CI 5.491–111.80) rests on a very small number and the authors flag it as exploratory. The overall cut-off of 24 has a sensitivity of 90.7%, so a non-trivial number of patients who do walk independently will score below it.
Summarised by AI from the full paper, without a clinician reviewing it. Check it against the source before it changes what you do. Read it on PubMed →