Table 1

Longitudinal Data for “Mr. Smith” (WQL = 12.63 prospective, = 5.10 retrospective)
1
 2
 3
 4
 5
 6
 7
Weeks after Baseline
 Weeks Before Death
 QOL
 QOLt
 QOLtd
 QOLtdi
 QOLback
0
 -18
 10
 .93
 .93
 .93
 10.00
1
 -17
 9
 .87
 .87
 .87
 9.00
2
 -16
 .

.92
 9.82
3
 -15
 9
 .87
 .87
 .87
 9.00
4
 -14
 9
 .87
 .87
 .87
 9.00
5
 -13
 .

.86
 8.77
6
 -12
 9
 .87
 .87
 .87
 9.00
7
 -11
 8
 .78
 .78
 .78
 8.00
8
 -10
 9
 .87
 .87
 .87
 9.00
9
 -9
 .

.74
 7.71
10
 -8
 8
 .78
 .78
 .78
 8.00
11
 -7
 .

.67
 7.20
12
 -6
 6
 .48
 .48
 .48
 6.00
13
 -5
 7
 .64
 .64
 .64
 7.00
14
 -4
 .

.54
 6.36
15
 -3
 7
 .64
 .64
 .64
 7.00
16
 -2
 .

.41
 5.63
17
 -1
 .

.33
 5.13
18
 0
 .

.23
 4.41
19
 .
 .
 .
 0
 0
 d
20
 .
 .
 .
 0
 0
 d
21
 .
 .
 .
 0
 0
 d
.
.
 .
 0
 0
 d
.
.
 .
 0
 0
 d
137
 .
 .
 .
 0
 0
 d
138
 .
 .
 .
 .
 .
 .
.
 .
 .
 .
 .
 .
 .
150
 .
 .
 .
 .
 .
 .

1 Weeks after enrollment (0 is person's baseline)

2 Weeks before death. (0 is the week the person died)

3 QOL: Observed QOL data

4 QOLt: QOL transformed to the probability of having good QOL next week given current QOL

5 QOLtd: QOLt with the weeks when person was dead set to 0. Weeks 138–150 are missing because Mr. Smith enrolled only 137 weeks before the end of current data collection.

6 QOLtdi: QOLtd with missing data imputed. Unknown for weeks 117–131 (for Mr. Smith).

7 QOLtdi transformed back to the original categories. (identical to observed QOL values, imputed values may need to be rounded).

Diehr et al. Health and Quality of Life Outcomes 2007 5:51   doi:10.1186/1477-7525-5-51