Nineteen individuals surviving the first COVID-19 wave accessed our clinic and consented to participate. The participants (male, 8) were 35–94 years old (median, 76 year). Time between hospitalization and healing/end of the isolation ranged from 16 to 129 days (median, 44 days). Frequent comorbidities were hypertension, hypercholesterolemia, heart and circulation disorders, and others (e.g., COPD, disorders of the prostate or thyroid glands). When the infection occurred, two individuals were admitted to the hospital for mild or severe stroke, and one for a femur fracture. All had been hospitalized in acute inpatient wards (intensive care unit, 4). Fifteen of them were transferred to low care wards, and four were discharged and isolated in non-hospital facilities (COVID-19 hotel, 2; home, 2). The individuals mostly accessed the baseline assessment session walking independently (alone, 9; accompanied, 8); two of them attended the clinic in a wheelchair.
According to the Pfeiffer short portable mental status questionnaire, three individuals had severe impairment, intellectual functioning was otherwise intact. The informed consent was given by the daughters in cases suffering from severe aphasia or intellectual impairment.
Individuals attended 1–15 sessions (median, 10) in 1–63 days (median. 42), and were monitored for 180–252 days (median, 223). The TUG test was not administered to individuals unable to walk independently and to the 94-year-old lady who reported high perceived exertion at rest (Borg Scale CR10, 7). The EuroQOL was not administered to the individual with severe aphasia. All but one individual were evaluated at T1; a participant felt recovered and withdrew from the intervention after the 3rd session. Two other individuals refused to participate in the T2 and T3 follow-up (did not like to evoke the illness or worsening of health condition). Telephone follow-up (i.e., assessing BI, PSFS, and EuroQOL) was administered to three individuals at T2, and four at T3.
At baseline about half of the individuals were independent in BADL, one quarter showed a slight dependency, whereas the remaining ones presented moderate to total dependency (Fig. 1). TUG and SPPB scores were 7.62–32.33 seconds (median, 12.42) and 0–12 points (median, 8), respectively (Fig. 1). Seven out of 16 individuals were at fall risk (i.e., TUG > 13.5 seconds) [10], and all but two presented mobility disability (i.e., SPPB score ≤ 10) [11]. Individuals mostly reported walking (quality and quantity) as the important activity they were having difficulty with because of their past condition, other activities frequently identified were housekeeping and climbing/descending stairs; PSFS score was 0–7 (median, 4). The EuroQOL-index showed a perceived health status in line with the population norms values (0.18-1.00; median, 0.87), whereas the EuroQOL-VAS scores were slightly below the average (50–85; median, 59) [13] (Fig. 1).
Table 1
Variable | N | | T0 | | T1 | | T2 | T3 | p value |
BI | 19 | | 98 (90 to100) | | 100 (97 to 100) | | 100 (96 to 100) | 100 (95 to 100) | .008a |
TUG | 16 | | 12.43 (10.03 to 16.40) | | 10.66 (8.43 to 14.66) | | 9.81 (8.22 to 15.29) | 8.99 (7.95 to 13.33) | .006a |
At fall risk | | | 7 (44%, 23–67) | | 4 (25%, 10–49) | | 5 (31%, 14–56) | 4 (25%, 10–49) | |
SPPB | 19 | | 8 (5 to 9) | | 9 (6 to 11) | | 11 (4 to 12) | 10 (4 to 12) | .002a |
Mobility disability | | | 17 (89%, 69–97) | | 12 (63%, 41–81) | | 9 (47%, 27–68) | 11 (58%, 36–77) | |
PSFS | 19 | | 4.0 (3.0 to 6.0) | | 7.0 (5.0 to 8.0) | | 7.0 (3.0 to 8.0) | 7.5 (5.0 to 8.5) | < .001a |
EuroQOl 5D index | 17 | | .87 (.80-.92) | | .90 (.80 to .92) | | .90 (.85 to 1.00) | .90 (.82 to 1.00) | .014a |
EuroQOl 5D VAS | 18 | | 59.0 (50.0 to 70.0) | | 75.0 (67.5 to 80.0) | | 80.0 (67.5 to 85.0) | 81.5 (57.8 to 90.0) | < .001a |
POST HOC b | | | Median of the differences (95%CI) c | | | | |
TUG | 16 | | T3 vs T0 | | -2.37 (-3.52 to -1.90) | | | | .004d |
SPPB | 19 | | T2 vs T0 | | 1.50 (0.50 to 2.50) | | | | .023d |
PSFS | 19 | | T1 vs T0 | | 2.0 (1.0 to 3.0) | | | | .050d |
| | | T2 vs T0 | | 2.0 (1.0 to 3.0) | | | | .012d |
| | | T3 vs T0 | | 2.3 (1.0 to 3.5) | | | | .002d |
EuroQOl 5D VAS | 18 | | T2 vs T0 | | 15.0 (9.0 to 24.0) | | | | .014d |
| | | T3 vs T0 | | 15.0 (7.5 to 23.0) | | | | .002d |
a Friedman tests. b Only significant results. c Hodges-Lehmann estimator. d Dunn tests with Bonferroni correction. Data are absolute frequencies, or median (1st to 3rd quartile), or absolute frequencies (relative frequencies, 95% confidence interval). Abbreviations: BI = Barthel Index; PSFS = Patient Specific Functioning Scale; SPPB = Short Physical Performance Battery test; TUG = Timed Up and Go test; VAS = Visual Analogue Scale. |
T0, T1, T2, and T3 assessments were performed 1–83 (median, 28.5), 40–106 (median, 77.5), 103–170 (median, 140), and 222–286 (median, 255) days after end of the isolation, respectively. According to intention-to-treat analysis, Friedman tests showed positive differences between the assessments (Table 1). Post-hoc analysis did not show differences among assessments in BI and EuroQOL-index; positive differences were evident for the TUG (T3 vs T0), SPPB (T2 vs T0), PSFS (T1, T2, and T3 vs T0), and EuroQOL-VAS (T2 and T3 vs T0) values (Table 1).
We observed variations in performance greater or smaller than the absolute MDC value. Compared to baseline SPPB values, five (26%, 95%CI 12–49), ten (53%, 95%CI 32–73), and seven (37%, 95%CI 19–59) individuals showed an improvement greater than the absolute MDC value at T1, T2, and T3, respectively, whereas one individual (5%, 95%CI 1–25) showed worsening at T2 and T3. Compared to baseline PSFS values, six (32%, 95%CI 15–54), five (26%, 95%CI 12–49), and eight (43%, 95%CI 42–64) individuals showed improvement at T1, T2, and T3, respectively, whereas one individual (5%, 95%CI 1–25) showed worsening at T2 (Fig. 2).