A total of 354 patients were randomised, with primary outcome assessment available for 346 patients (98%) (Figure 2). In both the intervention (n=170) and control group (n=176) the majority were female (67%, 62%), married/de facto (70%, 67%), had an informal carer (83%, 86%) and were on a 21-day cycle treatment protocol
Table 1 Characteristics of participants at baseline by treatment group
|
Intervention
N = 170
n (%)
|
Control
N = 176
n (%)
|
Gender
Male
Female
|
56 (32.9)
114 (67.1)
|
67 (38.1)
109 (61.9)
|
Age Mean (SD)
(range)
|
59.3 (12.7)
24 to 87
|
59.4 (13.6)
30 to 85
|
Marital status
Never married
Married/de facto
Separated/divorced
Widowed
Missing
|
28 (16.5)
119 (70.0)
15 ( 8.8)
6 (3.5)
2 (1.2)
|
26 (14.8)
117 (66.5)
15 (8.5)
14 (8.0)
4 (2.3)
|
Informal carer
Yes
No
Missing
|
141 (82.9)
21 (12.4)
8 (4.7)
|
151 (85.8)
17 (9.7)
8 (4.5)
|
Type of cancer
Breast
Colorectal
Genito-urinary
Gynaecological
Lung
Upper gastrointestinal
Other
|
56 (32.9)
20 (11.8)
18 (10.6)
33 (19.4)
21 (12.4)
19 (11.2)
3a (1.8)
|
41 (23.3)
30 (17.0)
15 (8.5)
24 (13.6)
20 (11.4)
31 (17.6)
15b (8.5)
|
Cancer stage
I
II
III
IV
Missing
|
7 (4.1)
25 (14.7)
46 (27.1)
79 (46.5)
13 (7.6)
|
6 (3.4)
30 (17.0)
31 (17.6)
90 (51.1)
19 (10.8)
|
Cycle duration based on chemotherapy protocol
7 days
14 days
21 days
28 days
|
5 (2.9)
40 (23.5)
112 (65.9)
13 (7.6)
|
4 (2.3)
38 (21.6)
116 (65.9)
18 (10.2)
|
Intent of treatment
Curative
Palliative
Not specified
|
72 (42.4)
59 (34.7)
39 (22.9)
|
73 (41.5)
65 (36.9)
38 (21.6)
|
Number of co-morbidities
Mean (SD)
Min-Max
Median (IQR)
|
1.4 (1.7)
0–9
1.0 (0.0–2.0)
|
1.7 (1.6)
0–7
1.0 (0.0–3.0)
|
a 1 x Mesothelioma, 1 x Head and Neck, 1 x Sarcoma
b 4 x Mesothelioma, 3 x Peritoneal, 3 x Head and Neck, 2 x Metastatic adenocarcinoma, 1 x Sarcoma,
1 x Brain cancer, 1 x Unknown primary
In the intervention group 106 (62%) patients and 126 (72%) patients in the control group reported comorbidities, obtained from the EMR. The most frequently reported comorbidities for the intervention vs control group were hypertension (25%, 31%), hypercholesterolaemia (9%, 7%), diabetes (9%, 11%), coronary heart disease (5%, 7%) and depression (7%, 5%).
Cycle delivery and completion details are reported in Table 2. In both groups most participants started all 4 cycles (85.0%, 86.0%). Ten patients died during the study, 6 in the intervention group and 4 in the control group. In both groups, most participants completed their cycles with no treatment delay or delays <7 days (85%, 88%). Additional information about delays in treatment is in the Supplement, Resource 2.
Table 2 Cycle delivery, completion and delays of participants by treatment group
|
Intervention
N = 170
n (%)
|
Control
N = 176
n (%)
|
Cycles commenced
C1 only
C1 & C2
C1, C2 & C3
C1, C2, C3 & C4
|
4 (2.4)
11 (6.5)
10 (5.9)
145 (85.3)
|
7 (4.0)
8 (4.5)
10 (5.7)
151 (85.8)
|
Delay to treatment >7 days
No
Yes
|
144 (84.7)
26 (15.3)
|
155 (88.1)
21 (11.9)
|
Deceased
|
6 (3.5)
|
4 (2.3)
|
Mean (SD) time at risk (days)
Range
Median (IQR)
|
57.3 (16.4)
8 to 112
63 (42–63)
|
57.9 (16.2)
3 to 97
63 (47–64)
|
Primary outcome: unplanned presentations
One or more unplanned presentation was made by 53 patients (31%) in the intervention group and 62 patients (35%) in the control group, resulting in 66/170 (mean 0.39) unplanned presentations in the intervention group and 78/176 (0.44) in the control group (Figure 3). Using Poisson regression to adjust for time at risk and site, the relative reduction in the number of unplanned presentations in the intervention compared to the control group was 12% (95% CI: −25%, 37%; P=0.48) and in the secondary analysis 10% (95%CI: −28%, 36%; P=0.56), after also adjusting for age, carer, cancer type and cancer stage. In both groups, most participants who made an unplanned presentation did so within the first 7 days of a chemotherapy cycle, mainly due to chemotherapy side effects. Most presentations (51%) were made during cycle 1 (Table 3).
Table 3 Characteristics of unplanned presentations by treatment group
|
Intervention
N = 66/170
n (%)
|
Control
N = 78/176
n (%)
|
Days since start of chemotherapy and presentation
1-7 days
8-14 days
15-21 days
22-28 days
29-43 days
|
29 (43.9)
25 (37.9)
9 (13.6)
2 (3.0)
1 (1.5)
|
31 (39.7)
26 (33.3)
18 (23.1)
1 (1.3)
2 (2.6)
|
Presented during cycle
1
2
3
|
36 (54.6)
17 (25.8)
13 (19.7)
|
38 (48.7)
23 (29.5)
17 (21.8)
|
Reason for presentation
Chemotherapy side effects
Disease progression
Unrelated to chemotherapy or disease
Combination of reasons
|
38 (57.6)
13 (19.7)
5 (7.6)
10 (15.2)
|
43 (55.1)
16 (20.5)
6 (7.7)
13 (16.7)
|
Symptom at presentationa
Fever without febrile neutropenia
Febrile neutropenia
Nausea and vomiting
Abdominal pain Diarrhoea
Constipation
Shortness of breath
Lethargy/Fatigue
Cough/sore throat
Confusion
Dehydration
Anxiety
Chest pain
Headache
Collapse
|
14 (21.2)
11 (16.7)
11 (16.7)
10 (15.2)
5 (7.6)
6 (9.1)
6 (9.1)
6 (9.1)
5 (7.7)
4 (6.1)
2 (3.0)
3 (4.5)
5 (7.6)
3 (4.5)
4 (6.1)
|
13 (16.7)
12 (15.4)
19 (24.4)
11 (14.1)
9 (11.5)
5 (6.4)
7 (9.0)
6 (7.7)
12 (15.6)
2 (2.6)
5 (6.4)
4 (5.1)
4 (5.1)
4 (5.1)
1 (1.3)
|
Number of symptoms (Mean, SD)
Range
|
2.2 (1.3)
1–7
|
2.3 (1.42
1–8
|
a Numbers add up to more than 66 and 78 and percentages add up to more than 100 as some patients presented with multiple symptoms
In total, 62 reasons for making an unplanned presentation were identified, with patients in both groups presenting with an average of 2 symptoms. The presenting symptoms, aggregated across all 3 cycles, that were reported for at least 5% in either group are shown in Table 3. In both the intervention and control group, most presentations for fever occurred during the first 7 days following treatment (64%, 31%), for febrile neutropenia within 8–14 days (73%, 58%) and for nausea and vomiting within the first 7 days (64%, 53%).
Unplanned admissions
There were 37/170 (mean 0.22) and 49/176 (mean 0.27) unplanned admissions to hospital (Table 4). In a post hoc analysis using negative binomial regression to allow for time at risk, the relative reduction in admissions following an unplanned presentation was 17% (95%CI: −37%, 50%, P=0.46) in favour of the intervention. Symptoms that were diagnosed on admission in ≥5% of either group are reported in Table 4, aggregated across all cycles.
Table 4 Characteristics of all unplanned hospital admissions by treatment group
|
Intervention
N = 37/170
n (%)
|
Control
N = 49/176
n (%)
|
Reason for admissiona
Febrile neutropenia
Respiratory infection
Fever without febrile neutropenia
Abdominal pain
Nausea and vomiting
Diarrhoea
Constipation
Shortness of breath
Anaemia
Electrolyte imbalance
Dehydration
Chest pain
Collapse
Back pain
Decreased consciousness
Confusion
Bleeding
Thrombocytopenia
Tachycardia
High creatinine
Lethargy/Fatigue
General weakness
Urinary tract infection
Pleural effusions
|
14 (37.8)
8 (21.6)
5 (13.5)
6 (16.2)
6 (16.2)
5 (13.5)
5 (13.5)
5 (13.5)
5 (13.5)
3 (8.1)
2 (5.4)
2 (5.4)
2 (5.4)
2 (5.4)
2 (5.2)
2 (5.4)
2 (5.4)
2 (5.4)
2 (5.4)
2 (5.4)
1 (2.7)
1 (2.7)
0 (0.0)
0 (0.0)
|
12 (24.5)
7 (14.3)
13 (26.5)
6 (12.2)
11 (22.4)
7 (14.3)
6 (12.2)
2 (4.1)
3 (6.1)
6 (12.2)
4 (8.2)
4 (8.2)
0 (0.0)
0 (0.0)
1 (2.0)
1 (2.0)
1 (2.0)
0 (0.0)
0 (0.0)
0 (0.0)
3 (6.1)
3 (6.1)
5 (10.2)
4 (8.2)
|
Number of symptoms (Mean, SD)
Min-Max
|
3.1 (2.3)
1–11
|
2.9 (1.5)
1–8
|
Total length of stay days (Median, IQR)
|
3.0 (2.0–9.0)
|
3.0 (2.0–7.0)
|
a Numbers add up to more than 36 and 50 and percentages add up to more than 100 as some patients admitted with multiple symptoms
Secondary outcomes
Results of the primary analysis of the PROs for the 336 patients who survived to the start of cycle 4 are shown in Table 5. For the HADS, after adjusting for site and baseline score, the final anxiety score was reduced by slightly more in the intervention than the control group by 0.47 (95%CI: −0.28 to 1.22; P=0.22). Depression scores were increased by slightly less in the intervention than the control group, a difference of 0.57 (95%CI: −0.18 to 1.31; P=0.13). The total HADS score showed that distress scores were decreased by somewhat more in the intervention than the control group, a difference of 1.05 (95%CI: −0.25 to 2.35; P=0.12). Results for all HADS scales were similar in the secondary analysis, which also showed that final anxiety scores were higher by an average of 1.45 (95%CI: 0.32 to 2.59; P=0.01) among patients with a carer than those without.
For CBI-B, where higher scores indicate greater coping efficacy, the increase was greater in the intervention group by 4.3 (95%CI: 0.7 to 7.9; P=0.02). There was no evidence of any effect of the intervention on the EORTC QLQ-C30 summary score, for which higher scores indicate better quality of life. Scores were reduced by only 0.16 more in the intervention group (95%CI: −2.67, 3.00; P=0.9). (See Resource 3 for all 15 subscale results). Results were again similar in the secondary analyses of both CBI-B and QLQ-C30 summary scores. For CBI-B, patients without a carer had greater coping self-efficacy than those with by an average of 6.4 (95%CI: 1.0 to 11.8; P=0.02).
Table 5 Comparison of secondary outcomes between groups
|
Intervention (N=164)
|
Control (N=172)
|
Differencea at end
Intervention – Control
|
P-value
|
|
Baseline
|
End of Cycle 3
|
Change
|
Baseline
|
End of Cycle 3
|
Change
|
|
|
HADS
Anxiety
Depression
Total (distress)
|
7.24
4.31
11.55
|
5.54
4.81
10.35
|
−1.70
0.50
−1.20
|
6.68
4.39
11.07
|
5.67
5.42
11.10
|
−1.00
1.03
0.03
|
−0.47 (−1.22, 0.28)
−0.57 (−1.31, 0.18)
−1.05 (−2.35, 0.25)
|
0.22
0.13
0.12
|
CBI-B
|
97.2
|
102.5
|
5.3
|
99.3
|
99.5
|
0.2
|
4.3 (0.7, 7.9)
|
0.02
|
EORTC summary score
|
76.21
|
72.76
|
−3.46
|
76.03
|
72.83
|
−3.20
|
−0.16 (−3.00, 2.67)
|
0.91
|
a difference between groups at end of Cycle 3 is adjusted for site