Hospital staff identified 7,114 patients (T1=2770, T2=2082, T3=2262) from the admission records meeting study eligibility criteria. Ward nurses further excluded 2,556 patients they were unable to locate (T1=722, T2=291, T3=358), or who were unable to communicate or in too poor physical status (T1=526, T2=286, T3=373). Finally, 739 patients declined to participate (T1=389, T2=161, T3=189), leaving 3,819 (T1=1133, T2=1344, T3=1342) patients who gave consent for telephone follow-up (See Figure 2). Following discharge, 3,083 patients (T1=896, T2=1093, T3=1094) completed the survey for a response rate of 83.5%, 86.1%, and 83.8%, respectively.
Participants did not differ by socio-demographic characteristics across the three time points (See Table 1). However, self-reported health characteristics varied by time point. T1 participants were less likely to have prior admissions and comorbidities than T2 and T3 participants, while T3 participants reported better perceived health than T1 and T2 participants.
Table 1. Descriptive Statistics
Category
|
T1 (N=896)
|
T2 (N=1093)
|
T3 (N=1094)
|
|
|
N (%)
|
N (%)
|
N (%)
|
p-value1
|
Sex
|
|
|
|
|
|
|
0.19
|
Male
|
477 (53.24)
|
582 (53.25)
|
545 (49.82)
|
|
Female
|
419 (46.76)
|
511 (46.75)
|
549 (50.18)
|
|
|
|
|
|
|
|
|
|
Age
|
|
|
|
|
|
|
0.50
|
18 - 24
|
45 (5.02)
|
60 (5.49)
|
52 (4.75)
|
|
25 - 34
|
76 (8.48)
|
95 (8.69)
|
76 (6.95)
|
|
35 - 44
|
131 (14.62)
|
128 (11.71)
|
140 (12.80)
|
|
45 - 54
|
173 (19.31)
|
211 (19.30)
|
234 (21.39)
|
|
55 - 64
|
224 (25.00)
|
278 (25.43)
|
263 (24.04)
|
|
65 - 80
|
247 (27.57)
|
321 (29.37)
|
329 (30.07)
|
|
Median
|
56
|
56
|
56
|
0.78
|
|
|
|
|
|
|
|
|
Education Level
|
|
|
|
|
|
|
0.86
|
Primary or below
|
283 (31.62)
|
349 (32.02)
|
333 (30.75)
|
|
Secondary
|
422 (47.15)
|
521 (47.80)
|
535 (49.4)
|
|
Tertiary
|
190 (21.23)
|
220 (20.18)
|
215 (19.85)
|
|
|
|
|
|
|
|
|
|
Marital Status
|
|
|
|
|
|
|
0.86
|
Never married
|
182 (20.31)
|
236 (21.71)
|
215 (19.93)
|
|
Married
|
602 (67.19)
|
721 (66.33)
|
726 (67.28)
|
|
Widowed/Separated
|
112 (12.50)
|
130 (11.96)
|
138 (12.79)
|
|
|
|
|
|
|
|
|
|
Self-Perceived Health
|
|
|
|
|
|
|
0.00
|
Good/Very good/ Excellent
|
283 (31.76)
|
334 (30.64)
|
439 (40.72)
|
|
Fair / poor
|
608 (68.24)
|
756 (69.36)
|
639 (59.28)
|
|
|
|
|
|
|
|
|
|
Length of stay
|
|
|
|
|
|
|
|
Median (Range)
|
4 (2-46)
|
4 (2-52)
|
4 (2-60)
|
0.05
|
|
|
|
|
|
|
|
|
Prior admission in last 12 months
|
|
|
|
|
|
0.00
|
0
|
357 (39.84)
|
328 (30.01)
|
319 (29.16)
|
|
1
|
174 (19.42)
|
231 (21.13)
|
239 (21.85)
|
|
≥2
|
365 (40.74)
|
534 (48.86)
|
536 (48.99)
|
|
Median (Range)
|
1 (0-139)
|
1 (0-156)
|
1 (0-157)
|
0.00
|
|
|
|
|
|
|
|
|
Number of comorbidities
|
|
|
|
|
|
0.00
|
0
|
297 (33.15)
|
345 (31.56)
|
309 (28.24)
|
|
1
|
276 (30.80)
|
256 (23.42)
|
287 (26.23)
|
|
≥2
|
323 (36.05)
|
492 (45.01)
|
498 (45.52)
|
|
Median (Range)
|
1 (0-11)
|
1 (0-14)
|
1 (0-13)
|
0.00
|
1 Chi-square test was used to evaluate differences in groups by time point.
Patient experience scores
The hospital baseline PPE-15 scores were all significantly different from the Hong Kong public hospital average (p<.05) (Table 2). Hospital baseline scores for the ‘information and education,’ ‘physical comfort,’ and ‘continuity and transition’ domains were higher (worse), while ‘coordination of care’, ‘emotional support,’ ‘respect for patient preferences,’ and ‘involvement of family and friends’ domains were significantly lower (better) than the Hong Kong public hospital average.
Table 2. Hong Kong Public Hospital 2005 Average and Study Hospital PPE-15 Summary, Domain and Item Scores
PPE-15 Items
|
2005 Hong Kong Public Hospital 1 (N=2901)
|
T1 (N=896)
|
T2 (N=1093)
|
T3 (N=1094)
|
p-value 2
|
Weighted% (95%CI)
|
Mean (95%CI)
|
Mean (95%CI)
|
Mean (95%CI)
|
Information and education
|
37.6 (35.0 - 40.2)
|
40.0 (37.3 - 42.6)
|
28.0 (25.8 - 30.3)
|
28.3 (26.0 - 30.6)
|
0.000
|
Nurses’ answer to questions not clear
|
39.2 (36.4 - 42.1)
|
41.4 (38.2 - 44.6)
|
25.9 (23.3 - 28.5)
|
29.0 (26.3 - 31.7)
|
0.000
|
Doctors’ answers to questions not clear
|
34.8 (32.0 - 37.7)
|
38.5 (35.3 - 41.7)
|
30.2 (27.5 - 32.9)
|
27.6 (25.0 - 30.3)
|
0.000
|
Coordination of care
|
|
|
|
|
|
Staff gave conflicting information
|
30.8 (28.0 - 33.7)
|
30.3 (27.3 - 33.3)
|
24.3 (21.8 - 26.9)
|
17.7 (15.5 - 20.0)
|
0.000
|
Physical comfort
|
|
|
|
|
|
Staff did not do enough to control pain
|
22.7 (20.2 - 25.3)
|
37.5 (33.7 - 41.3)
|
30.6 (26.8 - 34.3)
|
29.0 (25.4 - 32.6)
|
0.003
|
Emotional support
|
51.6 (49.3 - 53.9)
|
46.7 (44.2 - 49.2)
|
36.7 (34.5 - 38.8)
|
32.1 (30.0 - 34.1)
|
0.000
|
Nurses did not discuss anxiety or fears
|
53.9 (50.9 - 56.9)
|
49.3 (46.0 - 52.6)
|
37.8 (34.9 - 40.7)
|
36.9 (34.0 - 39.8)
|
0.000
|
Doctor did not discuss anxiety or fears
|
49.6 (46.6 - 52.6)
|
41.1 (37.8 - 44.3)
|
28.4 (25.7 - 31.1)
|
30.1 (27.4 - 32.8)
|
0.000
|
Difficult to find someone to talk to about concerns
|
51.3 (48.3 - 54.3)
|
49.7 (46.6 - 52.9)
|
43.6 (40.7 - 46.6)
|
29.0 (26.3 - 31.7)
|
0.000
|
Respect for patient preferences
|
51.9 (50.3 - 53.5)
|
38.9 (37.0 - 40.8)
|
31.3 (29.6 - 32.8)
|
20.5 (19.0 – 22.1)
|
0.000
|
Not sufficiently involved in decisions about treatment and care
|
89.0 (86.8 - 91.0)
|
65.5 (62.4 - 68.6)
|
55.0 (52.0 - 57.9)
|
32.6 (29.8 - 35.4)
|
0.000
|
Not always treated with respect and dignity
|
37.7 (34.8 - 40.6)
|
32.2 (29.1 - 35.2)
|
25.4 (22.8 - 28.0)
|
20.9 (18.5 - 23.3)
|
0.000
|
Doctors sometimes talked as if I was not there
|
30.0 (27.3 - 32.8)
|
19.0 (16.4 - 21.6)
|
13.5 (11.4 - 15.5)
|
7.8 (6.2 - 9.4)
|
0.000
|
Involvement of family and friends
|
48.6 (46.1 - 51.2)
|
46.6 (44.0 – 49.3)
|
35.2 (32.9 - 37.4)
|
34.3 (32.3 - 36.4)
|
0.000
|
Family did not get opportunity to talk to doctor
|
50.7 (47.7 - 53.6)
|
39.7 (36.5 - 42.9)
|
23.6 (21.1 - 26.1)
|
21.7 (19.2 - 24.1)
|
0.000
|
Family not given information needed to help recovery
|
46.4 (43.5 - 49.4)
|
53.4 (50.1 - 56.7)
|
46.6 (43.6 - 49.5)
|
47.5 (44.5 - 50.5)
|
0.005
|
Continuity and transition
|
28.8 (26.6 - 31.1)
|
43.3 (41.0 – 45.5)
|
42.2 (40.2 – 44.2)
|
36.3 (34.4 – 38.1)
|
0.000
|
Not told about danger signals to look for at home
|
30.0 (27.3 - 32.8)
|
56.5 (53.4 - 59.7)
|
50.6 (47.7 - 53.6)
|
49.7 (46.7 - 52.7)
|
0.006
|
Purpose of medicines not explained
|
28.5 (25.9 - 31.3)
|
21.6 (18.9 - 24.3)
|
19.4 (17.1 - 21.8)
|
11.4 (9.5 - 13.3)
|
0.000
|
Not told about medication side effects
|
24.3 (21.7 - 27.0)
|
54.9 (51.3 - 58.4)
|
60.3 (57.2 - 63.4)
|
51.2 (47.9 - 54.4)
|
0.000
|
Summary Score
|
41.8 (40.3 - 43.3)
|
41.9 (40.3 - 43.4)
|
34.1 (32.8 - 35.4)
|
29.1 (27.9 - 30.3)
|
0.000
|
|
|
|
|
|
|
|
Notes: Higher scores correspond with less satisfied patients. T1= Baseline 9 months pre-accreditation survey; T2=3 months post-accreditation survey; T3= 15 months post-accreditation survey
1 Chan SK, Wong IO, Tin KY, Fung A, Johnston JM, Leung GM. Satisfaction with inpatient care in a population-based Hong Kong Chinese sample. Quality & safety in health care. 2010;19(3):173-81.
2 ANOVA comparison of 3 time points
Overall, the PPE-15 summary scores and domain scores declined significantly (improved) over the study period (T1 to T3) (Table 2). Between T1 and T2, all domain scores declined except for the ‘continuity and transition’ domain, reflecting high scores for two items related to medication information. Between T2 and T3, the ‘continuity and transition’ domain showed significant improvement, while other domains continued to improve (i.e. ‘care coordination’, ‘emotional support’, ‘respect for patient preferences’ and ‘involvement of family and friends’) or hold steady (i.e. ‘information and education’ and ‘physical comfort’).
Table 3. Multiple Regression with Time Point (Post-accreditation) as Predictor of PPE-15 Domain and Summary Scores
|
|
T1 (N=896)
|
T2 (N=1093)
|
T3 (N=1094)
|
PPE-15 Domains
|
|
|
Parameter estimate (95% CI)
|
p-value
|
Parameter estimate (95% CI)
|
p-value
|
Information and education
|
|
Ref.
|
-11.9 (-15.3,-8.5)
|
0.000
|
-11.7 (-15.1,-8.2)
|
0.000
|
|
Adjusted
|
Ref.
|
-12.3 (-15.7,-8.8)
|
0.000
|
-11.2 (-14.7,-7.8)
|
0.000
|
Coordination of care
|
|
Ref.
|
-6.0 (-9.7,-2.2)
|
0.002
|
-12.6 (-16.3,-8.8)
|
0.000
|
|
Adjusted
|
Ref.
|
-6.4 (-10.2,-2.7)
|
0.001
|
-11.9 (-15.6,-8.1)
|
0.000
|
Physical comfort
|
|
Ref.
|
-6.9 (-12.2,-16.6)
|
0.010
|
-8.5 (-13.7,-3.3)
|
0.001
|
|
Adjusted
|
Ref.
|
-8.2 (-13.5,-2.9)
|
0.002
|
-8.8 (-14.1,-3.5)
|
0.001
|
Emotional support
|
|
Ref.
|
-10.0 (-13.3,-6.8)
|
0.000
|
-14.6 (-17.9,-11.4)
|
0.000
|
|
Adjusted
|
Ref.
|
-10.6 (-13.8,-7.4)
|
0.000
|
-14.0 (-17.3,-10.8)
|
0.000
|
Respect for patient preferences
|
|
Ref.
|
-7.7 (-10.1,-5.2)
|
0.000
|
-18.4 (-20.8,-16.0)
|
0.000
|
|
Adjusted
|
Ref.
|
-8.0 (-10.4,-5.7)
|
0.000
|
-18.6 (-21.0,-16.2)
|
0.000
|
Involvement of family and friends
|
|
Ref.
|
-11.5 (-14.8,-8.2)
|
0.000
|
-12.3 (-15.6,-9.0)
|
0.000
|
|
Adjusted
|
Ref.
|
-12.3 (-15.6,-8.9)
|
0.000
|
-11.6 (14.9,-8.2)
|
0.000
|
Continuity and transition
|
|
Ref.
|
-1.1 (-4.0,1.8)
|
0.470
|
-7.0 (-9.9,-4.1)
|
0.000
|
|
Adjusted
|
Ref.
|
-1.8 (-4.7,1.1)
|
0.218
|
-7.3 (-10.2,-4.4)
|
0.000
|
Summary score
|
|
Ref.
|
-7.8 (-9.7,-5.9)
|
0.000
|
-12.8 (-14.7,-10.9)
|
0.000
|
|
Adjusted
|
Ref.
|
-8.4 (-10.2,-6.5)
|
0.000
|
-12.6 (-14.5,-10.7)
|
0.000
|
Note: Covariates in the adjusted model include patient age, gender, self-reported education level, marital status, self-reported health status, length of stay, prior admissions and number of comorbidities, smoking habit, alcohol use, insurance and medical benefit statu
The multiple regression results (Table 3) confirmed the time point score comparisons. The T2 and T3 parameter estimates were declining (improving) for PPE-15 summary scores and all domain scores except for the ‘continuity and transition’ domain, for which the declining coefficient was only significant at T3. When adjusted for all covariates, the multiple regression parameter estimates were similar to the unadjusted estimates. Likewise, the sensitivity analyses confirmed similar findings among patients with multiple prior admissions and longer lengths of stay (>4 days) (data not shown). Patients with longer stays generally evaluated their experience more favourably (lower scores) than the full sample, while patients with multiple prior admissions indicated more problems. However, trends in patient experience scores among patients with longer stays and patients with multiple prior admissions were generally consistent with the full sample except for the ‘physical comfort’ domain, which while declining was not significant at T2 or T3 for patients with longer stays and T3 for patients with multiple prior admissions.