This prospective cohort study was conducted at Saint Paul’s Hospital Millennium Medical College (SPHMMC), a tertiary referral hospital in Addis Ababa, Ethiopia. The hospital’s Internal Medicine Department includes a neurology unit with a dedicated stroke ward and an on-site rehabilitation center staffed by specialized physiotherapists and nurses.

Data were collected from May 1 to October 30, 2023.

- Source population: All adult patients admitted to the medical wards of SPHMMC during the study period.

- Study population: All newly diagnosed stroke patients admitted to SPHMMC who met eligibility criteria.

- Inclusion: Adults aged ≥18 years with clinically and radiologically confirmed ischemic or hemorrhagic stroke.

- Exclusion: Patients with pre-stroke disability (modified Rankin Scale >2), those transferred from other facilities >48 hours after stroke onset, and patients with severe cognitive impairment assessed using the Mini-Mental State Examination (MMSE) score <18, precluding informed consent or reliable functional assessment.

A total of 152 consecutive eligible stroke patients were enrolled until the sample size was reached. Consecutive sampling minimized selection bias.

A pilot study validated the use of the Functional Independence Measure (FIM) in the local context. Data were collected using structured checklists covering sociodemographic variables, stroke characteristics, admission NIH Stroke Scale (NIHSS) scores, comorbidities, physiotherapy care details (enrolment, timing, intensity), and hospital service factors.

Eight internal medicine residents, trained in a two-hour session on standardized assessment and data recording, collected data. Supervisors ensured data completeness and consistency.

Functional recovery was assessed at discharge or three weeks post-stroke during outpatient follow-up. For patients hospitalized beyond three weeks, assessment was conducted as inpatients.

Independent variables: Age, sex, educational status; comorbidities (diabetes mellitus, hypertension, heart failure, dyslipidemia, chronic kidney disease); stroke type (ischemic or hemorrhagic), location (cortical or subcortical); stroke severity (NIHSS score); sensory loss; physiotherapy enrolment; physiotherapy intensity (total hours/session); timing of physiotherapy initiation.

Dependent variable: Functional recovery measured by change in FIM score at three weeks post-stroke.

Stroke severity: NIHSS score categorized as minor (1–4), moderate (5–15), and moderate to severe (>15).

Functional outcome: FIM total score (range 18–126), with higher scores indicating greater independence. A significant improvement was defined as a meaningful increase in FIM score leading to functional independence (>108).

Data were entered and analyzed using SPSS v26. Descriptive statistics summarized baseline characteristics. Associations between predictors and functional outcomes were assessed with chi-square tests for categorical variables and independent t-tests for continuous variables. Variables with p < 0.20 in bivariate analysis were included in multivariate binary logistic regression to identify independent predictors of functional improvement. Odds ratios (ORs) with 95% confidence intervals (CIs) were reported. Statistical significance was set at p < 0.05.

Missing data were minimal (<5%) and handled using listwise deletion.

The SPHMMC Institutional Review Board approved the study. Written informed consent was obtained from all participants or their caregivers. Confidentiality was maintained throughout.

Among 152 stroke patients, 72 (47.4%) were male and 80 (52.6%) female. The mean age was 63.0 ± 15.85 years (range 25–89). Sociodemographic characteristics and physiotherapy enrolment are summarized in Table 1.

Males had higher physiotherapy enrolment (44%) than females (35%). Enrolment varied by residence, highest among Oromia residents (57.1%) and lowest among Amhara (11.1%). Educational status influenced enrolment; inidividuals with a college degree had the highest participation rate (55%), while those with no formal education had the lowest (26.1%).

Patients without comorbidities were more likely to receive physiotherapy (63.3%) compared to those with comorbidities (23.9%). Ischemic stroke patients had higher enrolment (46.3%) than hemorrhagic stroke patients (31.4%). Stroke severity influenced enrolment: 47.2% of mild, 43.8% of moderate, and 25% of severe cases were enrolled. Absence of sensory impairment was associated with higher enrolment (44.7% vs. 31%). Mean admission FIM score was 67.66 ± 27.08, indicating moderate functional status (Table 2).

Only 39.5% (n=60) of patients received physiotherapy during hospitalization. Timing of initiation was delayed; only 6.6% started within 48 hours, 18.4% between 48 hours and 7 days, and 10.5% after 10 days. A majority (60.5%) did not receive inpatient physiotherapy. Hospital physiotherapy service availability was reported by 31.6%, but only 1.3% experienced regular service provision. Mean total physiotherapy duration was 1.50 ± 6.03 hours. Average inpatient stay was 10.99 ± 5.48 days (Table 3).

Medical treatments administered included anti-ischemic therapy (35.5%), antihypertensive agents (39.5%), anticoagulation (7.9%), and combined treatments (17.1%). In-hospital complications were common, with pneumonia affecting 40.3% of patients, followed by upper gastrointestinal bleeding (17.9%), venous thromboembolism (10.4%), acute kidney injury (9.0%), urinary tract infections (6.0%), and other complications (16.4%).

At three weeks, mean FIM score improved to 71.05 ± 30.97. The change ranged from –6 to +14 points. Forty percent of patients reached high functional independence (FIM > 80). Forty-three percent showed statistically significant functional improvement (Figure 1).

Mean-71.05, SD-30.968 N= 152

Multivariate logistic regression identified physiotherapy enrolment (AOR = 5.36; 95% CI, 4.50–10.75; p = 0.006) and physiotherapy intensity (AOR = 1.35; 95% CI, 1.17–1.56; p < 0.001) as significant positive predictors of functional recovery.

Older age was negatively associated with recovery (AOR = 0.94; 95% CI, 0.88–1.00; p = 0.048). Presence of comorbidities (AOR = 0.05; 95% CI, 0.004–0.53; p = 0.013) and in-hospital complications (AOR = 0.003; 95% CI, 0.000–0.049; p < 0.001) were strongly negatively associated.

Stroke type, stroke localization, sex, NIHSS score, timing of physiotherapy initiation, medical treatments, and length of stay were not significantly associated after adjustment (Tables 4 and 5).