- Email: [email protected]
Patient plus partner trial: A randomized controlled trial of 2 interventions to improve outcomes after an initial implantable cardioverter-defibrillator
Accepted Manuscript Patient Plus Partner Trial: A Randomized Control Trial of Two Interventions to Improve Outcomes following an Initial Implantable Cardioverter Defibrillator Cynthia M. Dougherty, ARNP, PhD, FAHA, FAAN, Elaine A. Thompson, RN, PhD, Peter J. Kudenchuk, MD, FHRS PII:
S1547-5271(18)31027-0
DOI:
10.1016/j.hrthm.2018.10.011
Reference:
HRTHM 7772
To appear in:
Heart Rhythm
Received Date: 22 February 2018
Please cite this article as: Dougherty CM, Thompson EA, Kudenchuk PJ, Patient Plus Partner Trial: A Randomized Control Trial of Two Interventions to Improve Outcomes following an Initial Implantable Cardioverter Defibrillator, Heart Rhythm (2018), doi: https://doi.org/10.1016/j.hrthm.2018.10.011. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
1 2
Patient Plus Partner Trial: A Randomized Control Trial of Two Interventions to Improve Outcomes following an Initial Implantable Cardioverter Defibrillator
3 Short title: P+P After an ICD
RI PT
4 5
Cynthia M. Dougherty ARNP, PhD, FAHA, FAAN Professor, University of Washington School of Nursing
SC
Elaine A. Thompson RN, PhD Professor Emeritus, University of Washington School of Nursing
Peter J. Kudenchuk, MD, FHRS Professor, University of Washington School of Medicine, Division of Cardiology
M AN U
6 7 8 9 10 11 12 13 14 15 16 17 18
TE D
19 Corresponding Author:
21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37
Cynthia M. Dougherty University of Washington School of Nursing 1959 NE Pacific Street, HSB T615A Box 357266 Seattle, WA 98195 [email protected] (206) 221-7927
AC C
EP
20
Word count: 4927
Funding: NIH, NHLBI R01HL086580-01A2 Clinicaltrials.gov: NCT01252615
The authors report no conflicts of interest.
1
ACCEPTED MANUSCRIPT
Abstract
M AN U
SC
RI PT
Background: Interventions to improve physical and psychological outcomes in recipients with an initial implantable cardioverter (ICD) and their intimate partners are largely unstudied, though likely to have a major impact on adjustment to the ICD and general well-being. Objective: To report the primary outcomes of the Patient Plus Partner randomized control trial. Methods: In a two-group (N=301) prospective RCT, we compared two social-cognitive-based intervention programs [Patient Plus Partner (P+P) and Patient Only (P only)], implemented following initial ICD implant. The patient intervention, consisting of educational materials, nursedelivered telephone coaching, videotape demonstrations, and access to a nurse via a 24/7 pager, was implemented in both groups. P+P also incorporated partner participation. The primary patient outcomes were symptoms and anxiety at 3 months. Other outcomes were physical function (SF-36, ICD shocks-patient); psychological adjustment (PHQ-9); relationship impact (DAS, OCBS-partner); self-efficacy and knowledge (SCA-SE, SCA-OE, KSA); and health care utilization (outpatient visits, hospitalizations), at hospital discharge, 1, 3, 6, and 12 months post-ICD. Results: For patients, P+P compared to P Only was more effective in improving symptoms (P=0.02), depression (P=0.006), self-efficacy (P=.02), outcome expectations (P=0.03), and knowledge (P=0.07). For partners, P+P was more effective in improving partner caregiver burden (P=0.002), self-efficacy (P=0.001) and ICD knowledge (P=0.04). Conclusion: An intervention that integrated the partner into the patient’s recovery following an ICD improved outcomes for both. Beyond survival benefits of the ICD, intervention programs designed to address both the patient and their partner living successfully with an ICD are needed and promising.
EP
TE D
Key Words: Sudden cardiac arrest, ICD, clinical trial, self-efficacy, telephone intervention, partner, anxiety, depression, health care use
AC C
38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69
2
ACCEPTED MANUSCRIPT
INTRODUCTION
70 71
Research points to the important reciprocal influences that patients and partners have on one another’s recovery experiences after a cardiac event [1]. However, interventions
73
designed to impact outcomes of both patients with an implantable cardioverter defibrillator (ICD)
74
and their intimate partners, have not been systematically tested. Given the family stress caused
75
by life-threatening illness and the ever-increasing number of ICDs being implanted worldwide
76
[2], interventions are needed that are timely, both patient and partner-relevant, and integrated in
77
a manner that improves patient recovery. In our past research, we demonstrated that intimate
78
partners of recent ICD recipients reported better physical health than the patient, but their health
79
declined during the first year post-ICD. Intimate partners also reported significantly higher levels
80
of anxiety and depression over the first 12-months, higher than patients themselves. Following
81
ICD implantation, partners consistently reported high caregiving demands and a corresponding
82
lower level of satisfaction with their relationship with the patient [3,4]. Partners also note that
83
important information and support was typically not available to them in outpatient settings.
SC
M AN U
TE D
84
RI PT
72
We previously reported outcomes of a patient-focused self-efficacy intervention following an initial ICD implant after successful resuscitation from sudden cardiac arrest [5,6]. At the
86
conclusion of the intervention, we interviewed patients to determine essential elements of the
87
intervention. Most study participants wanted more telephone contact with the nurse, and would
88
have liked an accompanying videotape to demonstrate important skills; many also desired more
89
psychological support, and requested an intervention to better support their intimate partners.
90
We incorporated these suggestions into our original patient only (P Only) intervention, and
91
created an expanded Patient Plus Partner (P+P) intervention [6]. We then designed and
92
conducted a prospective randomized clinical trial comparing outcomes between the two
93
interventions, one that focused strictly on the ICD recipient (P Only) and one that was expanded
94
to include the partner (P+P). We hypothesized that the expanded (P+P) intervention would
95
result in better physical and psychological outcomes for both patients and partners, compared to
AC C
EP
85
3
ACCEPTED MANUSCRIPT
96
the intervention that focused on the patient alone (P Only). Secondarily, we tested for
97
differences in outcomes by ICD indication (e.g., primary vs. secondary prevention). METHODS
98 Design
RI PT
99
ICD recipients were randomized using a prospective 2-group blocked clinical trial (RCT)
101
design. The strata were Charlson Co-morbidity score (<2 or >2) [10] and ICD indication (primary
102
vs. secondary prevention). The RCT compared two intervention groups: 1) P Only, and 2) P+P
103
during the first year post-ICD implant. Study outcomes were: a) physical functioning in the
104
patient with an ICD (patient concerns/symptoms, general health, ICD shocks) and in the partner
105
(partner concerns/symptoms, general health); b) psychological adjustment (anxiety and
106
depression) in both the patient and partner; c) relationship impact for the patient (dyadic
107
adjustment) and partner (dyadic adjustment, caregiver burden); d) self-efficacy for both the
108
patient and partner, and e) health care utilization (outpatient visits, hospitalizations) for both
109
patient and partner.
110
Sample and Setting
TE D
M AN U
SC
100
Study participants (N=301) were identified as first-time ICD recipients based on
112
systematic screening at 12 acute care institutions in Washington, Oregon, Georgia, and North
113
Carolina between 2010-2015. Identified patients and their partners were approached for study
114
participation while hospitalized for the initial ICD implant. Written consent from both were
115
required for study eligibility. Study procedures were approved by hospital-based IRBs and the
116
academic IRB providing study oversight. Inclusion criteria were: 1) first ICD implant for primary
117
or secondary prevention of sudden cardiac arrest (SCA); 2) ability to read, speak, and write
118
English; 3) access to telephone for 1 year after ICD implant; and 4) having an intimate partner
119
(spouse, lover, life partner or significant other) involved in a committed relationship with the
120
patient. Exclusion criteria were: 1) clinical co-morbidities that impaired cognitive or physical
121
function; 2) short BLESSED cognitive screening score > 6 [7]; 3) age < 21 years; 4) AUDIT-C
AC C
EP
111
4
ACCEPTED MANUSCRIPT
122
score > 4 for alcohol use [8]; and/or 5) ASSIST 2.0 score > 4 for daily non-medical use of illicit
123
substances [9].
124
Protocol and Interventions After baseline data were collected, subjects were randomized to one of the two
126
interventions (P Only or P+P). A computerized program generated a balanced randomization
127
taking into account ICD indication, age, gender, and Charlson co-morbidity index [10]. All
128
participants completed a battery of measurements at the same time periods, which included
129
baseline (hospital discharge), and 1, 3, 6, and 12 month follow-ups, after which they received
130
study-related intervention materials. All participants received usual care from their providers as
131
well as interim telephone calls from study staff every 3 months to collect health care utilization
132
information. A description of the study interventions and intervention fidelity is found in the
133
supplementary materials.
134
Outcome Measures
M AN U
SC
RI PT
125
Outcomes were assessed at baseline (hospital discharge), and at 1 (midway through the
136
intervention, 3 (intervention completion), 6, and 12 months. Outcome measures, summarized in
137
Table S1 (supplemental materials), were obtained for all patients and their respective partners
138
in both intervention groups.
139
Analysis
EP
Prior to the study, we estimated that a total of 300 patients would be needed to provide
AC C
140
TE D
135
141
80% power to detect group differences in the primary outcomes of physical functioning (PCA-
142
symptoms) and psychological adjustment (STAI-anxiety) at 3 months, adjusted for an
143
anticipated 15% dropout. All analyses were conducted using an intent-to-treat approach.
144
Initially, we reviewed data for their distributional properties (means, variances, kurtosis) and
145
potential outliers. Using chi square or t-tests, we tested for group equivalence in baseline
146
demographic and outcome measures. Group intervention effects were tested for change in the
147
primary outcomes from baseline to 3, 6 and 12 months using repeated analysis of co-variance
5
ACCEPTED MANUSCRIPT
(ANCOVA, SPSS 21.0, SPSS Inc, Chicago, IL), controlling for baseline outcome values, age,
149
sex, Charlson score, ejection fraction %, and ICD indication (primary vs. secondary) [24].
150
ANCOVA F statistic indicated if the means between the two groups differed significantly. In
151
conducting secondary comparisons, Bonferroni corrections were made to adjust for potential
152
Type I error by dividing the α value by the number of additional comparisons (α/N) made in each
153
outcome category by timeframe. Statistical significance was defined at α ≤ 0.02, 2-tailed.
154
Intervention effects were also examined by indication for ICD implantation (primary vs.
155
secondary prevention). RESULTS
157
M AN U
156
SC
RI PT
148
Participants
Over the five year study, 1051 patient-partner pairs were approached for study
159
participation, of these 861 were eligible (Figure 1). The patient who received an ICD was the
160
“index case” for study referral. Of the 861 eligible patients, 301 patient-partner pairs met
161
eligibility criteria and were randomized to one of the two study groups: 151 pairs to the P Only,
162
and 150 to the P+P intervention. Of these, 92% of patients in the P Only 95% of patients in the
163
P+P group completed 12-month assessments. The follow-up rate over 12 months for partners
164
was similar to that of patients in both study groups (91% of partners in P Only; 94% of partners
165
in P+P).
167
EP
Patients
AC C
166
TE D
158
Study patients in both P Only and P+P groups averaged 64.14+11.9 (range 26-93)
168
years, had a Charlson co-morbidity score of 2.29 ± 1.49 and a body mass index (BMI) of 29.57
169
± 6.2. The majority were male (73.8%), Caucasian (91.0%), had some college education
170
(50.8%), and were retired from work (43.2%) or disabled (15.6%), and received an ICD for
171
primary prevention of cardiac arrest (58.9%) or secondary prevention of cardiac arrest (40.2%).
172
At baseline enrollment, 37.5% of patients reported a sedentary lifestyle, 20.6% were taking anti-
6
ACCEPTED MANUSCRIPT
173
depressant medications, 11.0% were taking anti-anxiety medications, and 13% were current
174
smokers (Table 1 and Table S2 in supplemental materials). Partners
176
Study partners in both P Only and P+P groups averaged 62.46+12.4 (range 28-91)
RI PT
175
years, had a Charlson co-morbidity score of 0.72+1.08 and a BMI of 28.9 ± 6.5. The majority
178
were female (74.1%), Caucasian (88%), had some college education (39.9%), and were retired
179
from work (40.5%) or disabled (6%). At baseline enrollment, 31.3% of partners reported a
180
sedentary lifestyle, 25.6% were taking anti-depressant medications, 18.9% were taking anti-
181
anxiety medications, and 12.3% were current smokers. (Table 1, and Table S2 supplemental
182
materials).
M AN U
SC
177
By comparison, patients and their partners in P Only and P+P groups did not differ on
183
many baseline factors. However, patients were less likely to be actively employed and their
185
partners more likely to be taking antidepressant medications in the P Only than P+P group;
186
whereas ASSIST recreational drug use was higher among patients and partners in the P+P
187
group.
188
Intervention Fidelity
Over the 5 year course of the trial, the proportion of intervention elements delivered
EP
189
TE D
184
between treatment groups (98.5 +1.2% in P Only and 97.5+2.3% in P+P), and the number of
191
phone calls completed (98.5%+1.2 in P Only and 97.5%+2.3 in P+P) were similar. The
192
cumulative time spent on telephone calls was significantly higher for those in the P+P group
193
(mean+SD) (322+92 minutes vs 250+80 minutes, p=0.001), given that engaging both the
194
partner and patient required more time to deliver the intervention. Partners who were
195
randomized to P+P participated in an additional 291+77 cumulative minutes in the PTG
196
intervention; 85% of whom completed all of the planned telephone calls. Most patients (97%)
197
and partners (96%) were favorably impressed by the program and recommended that others
198
participate in a similar one, if offered. Variability in intervention delivery by nurses within the
AC C
190
7
ACCEPTED MANUSCRIPT
intervention groups did not differ significantly, and the effect of the nurse on patient outcomes
200
was not significant.
201
Outcomes
202
Physical Function
203
RI PT
199
Patient Physical Function. The frequency of self-reported PCA symptoms (primary outcome) fell significantly between the intervention groups (Figure 2A and Table S3,
205
supplemental materials) at 12 months, with patients in P+P reporting fewer symptoms than
206
those in P Only (P = 0.02). Fear of dying and distress reduced in both groups over time.
207
Patients in both intervention groups reported lower levels of physical (PCS) than mental health
208
(MCS) at study enrollment and throughout 12 months, with no statistically significant by group
209
differences noted over 12 months. The number of ICD shocks by intervention group did not
210
differ across time, ranging from 3 to 9% over the course of 12 months. Overall ICD shock rates
211
were low, perhaps related to changes in ICD programming standards that occurred during the
212
course of the study.
M AN U
TE D
213
SC
204
Partner Physical Function. As well, partners in both intervention groups reported lower levels of physical (PCS) vs. mental (MCS) health throughout the study, with no statistically
215
significant by group differences noted. The number of physical symptoms reported by partners
216
was lower than patients, and overall general health of partners was better than patients, from
217
enrollment to final follow-up (Table S4, supplemental materials). The number of self-reported
218
symptoms among partners did not differ significantly between the intervention groups over 12
219
months. ICD and health-related partner concerns declined in both intervention groups over time.
220
Psychological Adjustment
221
AC C
EP
214
Patient anxiety (STAI) declined over time (primary outcome), with no significant
222
differences observed between intervention groups (Table S3). However, for P+P vs. P Only,
223
patient depression and depression severity (Figure 2B), significant reductions (P = 0.006 and
224
0.01, respectively) were noted at 12 months. Conversely, partner anxiety and depression
8
ACCEPTED MANUSCRIPT
225
remained relatively stable across the 12-months, without significant differences between P+P
226
and P Only.
227
Relationship Impact Patient Relationship Impact. No statistically significant differences were observed over
RI PT
228
time for P+P vs. P Only with respect to overall dyadic adjustment or subscales of dyadic
230
adjustment, including satisfaction, cohesion, consensus, or affectionate expression among
231
study patients (Table S3). Nor were there significant group differences in patient sexual
232
concerns, which were highest at hospital discharge and declined only modestly over time.
SC
229
Partner Relationship Impact. Similar to patients, there were no statistically significant
234
intervention group differences among partners in overall or subscale measurement of dyadic
235
adjustment, or sexual concerns (Table S4). Partner sexual concerns were higher than patients
236
during the first 3 months post-ICD, and reduced to levels similar to those of patients at 6
237
months. For caregiving burden, however, partners in P+P compared to P Only reported
238
significantly lower levels of caregiver burden and time spent in caregiving (Figure 2D). There
239
were also trends toward reduction in caregiving difficulty reported by partners in the P+P
240
compared to P Only at 3, 6, and 12 months, (P = 0.06, 0.07, and 0.06, respectively).
241
Self-Efficacy
TE D
EP
242
M AN U
233
Patient Self-Efficacy. Self-efficacy was assessed using three indicators: self-efficacy for living with the ICD, outcome expectations after an ICD, and ICD knowledge. For specific
244
aspects of self-efficacy for living with an ICD measured at 12 months, for P+P vs. P Only there
245
were significantly higher levels in ICD function (P = 0.01), managing physical changes caused
246
by the ICD (F=10.86, P = 0.001), relationship management related to the ICD (F=4.66, P =
247
0.03), and a trend for ICD safety (F=2.90, P= 0.09). For outcome expectations, a trend toward
248
improvement for patients in P+P vs. P Only was observed at 3 months (F=3.34, P = 0.07),
249
which was more pronounced at 6 months (F=5.06, P = 0.03), but declined by 12 months
AC C
243
9
ACCEPTED MANUSCRIPT
250
(F=0.36, P = 0.55). Patient ICD knowledge did not significantly differ over time for P+P vs. P
251
Only, but was slightly higher in P+P at most points in time. Partner Self-Efficacy. For partner self-efficacy for living with the ICD, the P+P partner
253
group reported significantly higher self-efficacy at all data collection points across 12 months
254
(Figure 2C; Table S4). For outcome expectations, there were no significant group differences
255
across time. Partner ICD knowledge was significantly higher in P+P vs. P Only at all time points
256
over 12 months.
257
Health Care Utilization
SC
Patient Health Care Utilization. No statistically significant group differences in the
M AN U
258
RI PT
252
frequency of patient hospitalization were noted over 1 year (Table S3). The patient
260
hospitalization rate in both groups was 7.9% in the first month post-ICD, 20.6% at 1-3 months,
261
23.3% at 3-6 months, and 36.9% at 6-12 months. The average number of patient hospitalization
262
days did not differ significantly by treatment group over 12 months, with total hospital days in
263
both groups ranging from 2-6 days. Patients were hospitalized during the first year post-ICD for
264
cardiac issues such as chest pain, heart failure exacerbation, cardiac arrhythmias, or ICD
265
related reasons. However, patients in P+P had fewer outpatient visits over the course of the trial
266
than those in P Only (2.54 vs. 3.09, P = 0.03; data not shown), which were most often ICD-
267
related for both groups.
EP
Partner Health Care Utilization. No statistically significant differences in the frequency of
AC C
268
TE D
259
269
partner hospitalizations were found between treatment groups during the follow-up year (Table
270
S4). The hospitalization rate among partners in both groups was 2.7% during the first month,
271
3.3% from 1-3 months, 11.3% from 3-6 months, and 19.3% from 6-12 months, primarily for non-
272
cardiac surgical procedures. The average number of hospitalization days in partners did not
273
differ significantly by group over the 12 months, ranging from 2-5 days. No significant group
274
differences were observed in the number of outpatient visits by partners over 12 months. On
10
ACCEPTED MANUSCRIPT
275
average, partners in both groups had 1 outpatient visit in each 3-month interval across the 12
276
months.
277
ICD Indication: Primary vs. Secondary Prevention We conducted exploratory analyses to determine the effects of ICD indication (primary
RI PT
278
vs. secondary prevention) on differential benefit from the interventions (data not shown). There
280
were marked baseline differences in the health of patients receiving an ICD for primary vs.
281
secondary prevention reflected in higher Charlson co-morbidity scores (t = 2.88, P = 0.004),
282
higher BMI (t = 2.17, P = 0.03), and lower ejection fraction (t = 8.64, P = 0.001) in the primary
283
prevention group. Outcomes at 3 months showed that patients receiving an ICD for secondary
284
prevention, had greater reductions in symptoms (PCA, F = 7.70, P = 0.006), better physical
285
health (SF-36 PCS, F = 8.91, P = 0.001), and greater ICD knowledge (F = 5.02, P = 0.02).
286
These changes were sustained at 12 months. Patients who received an ICD for primary
287
prevention experienced more hospitalizations (F = 5.48, P = 0.04) over the first 3 months, but
288
this difference was no longer apparent by 12 months. By comparison, ICD indication did not
289
appear to have a robust or consistent influence on partner responses over the course of the
290
trial.
292
M AN U
TE D
DISCUSSION
EP
291
SC
279
The health impact of an ICD reaches far beyond rhythm management alone, extending to the functional and psychological aspects of daily living for both patients and their intimate
294
partners. The importance of this RCT lies in being the first to establish such benefits from an
295
intervention directed at both the patient and partner. We demonstrated that a P+P intervention
296
is superior to a P Only intervention in reducing patient symptoms and depression, while
297
improving self-confidence and ICD knowledge over time. The P+P intervention also proved to
298
be superior to a P Only approach for the partner in reducing caregiver burden, while
299
concurrently improving self-confidence and knowledge.
AC C
293
11
ACCEPTED MANUSCRIPT
300
The intervention framework was based on Social Cognitive Theory [11], which is well suited for patients undergoing initial ICD implant and their partners. The ICD implant recovery
302
period requires acquisition of new knowledge, behavioral skills and confidence to promote living
303
successfully with an implanted device that is known to impact quality of life. A central
304
component of this framework is self-efficacy, the belief in one’s capacity to engage in new
305
behaviors in new situations. Enhancing self-efficacy as a means to facilitate health behavior
306
change in CV disease has been studied extensively, demonstrating that self-efficacy influences
307
cardiac-related lifestyle changes [25], motivation, and emotional reactions to stressful situations
308
[26]. For patients recovering from acute myocardial infarction (MI) and heart surgery, self-
309
efficacy have been positively associated with adherence to exercise programs [27-30].
310
Specifically, with coaching from health care providers, self-efficacy for walking and running
311
generalized to activities such as weight lifting and sexual intercourse [31]. Higher self-efficacy in
312
partners has been shown to predict long-term survival in patients with heart failure [32].
SC
M AN U
The content of the present intervention was tested earlier by our research group in
TE D
313
RI PT
301
patients undergoing initial ICD implant for secondary prevention [5,6]. The current study
315
included patients who had received an ICD for secondary as well as primary prevention.
316
Comparing study outcomes with our previous work [5,6], we note the following similarities in
317
both studies: 1) patients report significant reductions in symptoms and improved self-efficacy
318
and knowledge over 12 months, and 2) patient outcomes improved the most between hospital
319
discharge and 3 months, and then stabilized over the remainder of the year.
AC C
320
EP
314
Analyses based on primary vs. secondary indication demonstrated that reasons for
321
receiving an ICD were associated with differential response to the intervention in patients. The
322
patient group receiving an ICD for secondary prevention demonstrated a stronger response to
323
the intervention, whereas there were few differences in partner responses to the intervention
324
based on ICD indication.
325
Strengths and Limitations
12
ACCEPTED MANUSCRIPT
The strengths of this study were its randomized design, careful and detailed attention to
327
intervention fidelity, and low drop-out rates in both patient and partner groups. Few intervention
328
programs for patients who get an initial ICD have been developed and tested, and until now
329
none have included the partner. The interventions were delivered in the outpatient setting using
330
telephone delivery, without the need for costly in person clinic visits. The interventions are being
331
further individualized and developed for implementation into routine clinical care.
Though randomized, this trial was not blinded. Study outcomes were analyzed without
SC
332
RI PT
326
knowledge of the subject’s group assignment. The study design did not include a true patient
334
control group for comparison against a patient intervention group. However, the trial did include
335
a partner usual care group. Some study outcomes relied on self-report. In anticipation of this
336
limitation, individuals were excluded from study participation if they had cognitive deficits, were
337
in extremely poor physical health after ICD implant, or lacked a intimate partner who was willing
338
to participate. Study results pertain principally this population. Finally, while patients in the P
339
Only group were asked not to share intervention content with their partners, we had little control
340
over partner exposure to the written intervention content. However, partner participation in the
341
telephone portion of the P Only intervention was monitored to ensure that partners did not
342
participate in the weekly calls.
EP
TE D
M AN U
333
343
A telephone delivered self-efficacy based intervention that included both the patient and
AC C
344
CONCLUSIONS
345
the intimate partner following implantation of an initial ICD, resulted in significant reductions in
346
patient symptoms and depression, with higher self-efficacy, outcome expectations, and
347
knowledge over the P only intervention. Benefits to the partner included a reduction in caregiver
348
burden and improved self-efficacy and knowledge to care successfully for the patient with an
349
ICD. Post-hospital based intervention programs established for those with an ICD should
350
include both the patient and their intimate partner in education, support, and home care
351
management.
13
ACCEPTED MANUSCRIPT
References
352
354 355
1. Van Horn E, Fleury J, Moore S. Family interventions during the trajectory of recovery from cardiac event: an integrative literature review. Heart Lung 2002; 31:186-98. 2. Munir MB, Alqahtani F, Aljohani S, Bhirud A, Modi S, Alkhouli M. Trends and predictors of
RI PT
353
implantable cardioverter defibrillator implantation after sudden cardiac arrest: Insight from the
357
national inpatient sample. Pacing Clin Electrophysiol. 2018 Jan 9. doi: 10.1111/pace.13274.
358
[Epub ahead of print].
359
SC
356
3. Dougherty CM, Fairbanks AM, Eaton LH, Morrison MK, Kim MS, Thompson EA. Comparison of patient and partner quality of life and health outcomes in the first year after an
361
implantable cardioverter defibrillator (ICD). J Behav Med 2016, 39:94-106.
M AN U
360
362
4. Dougherty CM, Thompson EA. Intimate partner physical and mental health after sudden
363
cardiac arrest and receipt of an implantable cardioverter defibrillator. Res Nursing Health
364
2009; 32:432-442.
366 367
5. Dougherty CM, Thompson EA, Lewis FM. Long term outcomes of a nursing intervention
TE D
365
after an ICD. Pacing Clin Electrophysiol 2005; 28:1157-67. 6. Dougherty CM, Lewis FM, Thompson ET, Baer, JD, Kim W. Short term efficacy of a telephone intervention after ICD implantation. Pacing Clin Electrophysiol, 2004; 27:1594-
369
1602.
7. Blessed G, Tomlinson BE, Roth M. The association between quantitative measures of
AC C
370
EP
368
371
dementia and of senile change in the cerebral grey matter of elderly participants. British J
372
Psychia 1986;512:797-811.
373 374 375 376 377
8. Bush K, Kivlahan DR, McDonell MB, Fihn SD, Bradley KA. The AUDIT Alcohol Consumption Questions (AUDIT-C). Arch Intern Med 1998;158:1789-1795. 9. WHO Assist Working Group. The Alcohol, smoking, and substance involvement screening test (ASSIST): Development, reliability, and feasibility. Addict 2002; 97:1183-94. 10. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic
14
ACCEPTED MANUSCRIPT
378
comorbidity in longitudinal studies: Development and validation. J Chronic Dis 1987;40:373-
379
383.
381 382
11. Bandura A. Self-efficacy: Toward a unifying theory of behavioral change. Psych Rev 1977; 84:191-215.
RI PT
380
12. Dougherty CM, Benoliel JQ, Bellin CM. Domains of nursing intervention after sudden
cardiac arrest and automatic internal cardioverter defibrillator implantation. Heart Lung 2000;
384
29:79-86.
SC
383
13. Dougherty CM, Pyper GP, Benoliel JQ. Domains for nursing intervention for intimate
386
partners of sudden cardiac arrest survivors. J of Cardiovac Nurs 2004; 19:21-31.
387
M AN U
385
14. Dougherty CM, Thompson EA, Kudenchuk PJ. Development and Testing of an Intervention
388
to Improve Outcomes for Partners following Receipt of an Implantable Cardioverter
389
Defibrillator (ICD) in the Patient. Adv Nurs Sci 2012, 35:359-377. 15. Jenkins LS. Quality of life in patients of sudden cardiac death entering treatment. In Dunbar
391
SB, Ellenbogen KA, Epstein AE. Sudden Cardiac Death: Past, Present, and Future. Armonk:
392
NY: Futura Publishing;1997,289-305.
395 396 397 398 399 400 401 402
1987; 40:473-80.
EP
394
16. Ware JE. Standards for validating health measures: Definitions and content. J Chronic Dis
17. Spielberger C, Gorsuch R, Lushene R. Manual for the state-trait anxiety inventory (selfevaluation questionnaire). Palo Alto, CA: Consulting Psychologists, 1970.
AC C
393
TE D
390
18. Kroenke K, Spitzer RL, Williams JB. The PHQ-9: validity of a brief depression severity measure. J Gen Int Med 2001;16:606-613. 19. Steinke EE, Gill-Hopple K, Valdez D, Wooster M. Sexual concerns and educational needs after an implantable cardioverter defibrillator. Heart Lung 2005: 34:299-308. 20. Steinke EE. Sexual concerns of patients and partners after an implantable cardioverter defibrillator. Dimens Crit Care Nurs 2003; 22:89-96.
15
ACCEPTED MANUSCRIPT
403 404 405
21. Spanier G.B. Assessing dyadic adjustment: New Scales for assessing the quality of marriage and similar dyads. J Marriage Fam 1976; 38:15-28. 22. Bakas T, Austin JK, Jessup SL, Williams LS, Oberst MT. Time and difficulty of tasks provided by family caregivers of stroke survivors. J Neurosci Nurs 2004, 36: 95-105.
407
23. Dougherty CM, Thompson EA, Johnston SA. Reliability and validity of the self-efficacy
RI PT
406
408
expectations and outcome expectations after ICD implantation scales. Appl Nurs Res 2007;
409
20:116-124.
Saddle River, NJ: Pearson/ Prentice Hall, 2004.
SC
411
24. Kepple G, Wickens TD. Design and analysis: A researcher’s handbook (4th Ed). Upper
M AN U
410
412
25. Bandura A. Self-efficacy mechanism in human agency. Am Psych 1982; 37:122-147.
413
26. Izawa KP, Watanabe S, Omiya K, Hirano Y, Oka K, Osada N, Iijima S. Effect of the self-
414
monitoring approach on exercise maintenance during cardiac rehabilitation: A randomized
415
controlled trial. Am J Phys Med Rehab 2005; 84:313-21.
27. Taylor CB, Bandura A, Ewart CK, Miller NH, DeBusk RF. Exercise testing to enhance wives
TE D
416 417
confidence in their husbands cardiac capability soon after clinically uncomplicated acute
418
myocardial infarction. Am J Card 1985; 55:635-38. 28. Sniehotta FF, Scholz U, Schwarzer R, Fuhrmann B, Kiwus U, Voller H. Long term effects of
EP
419
two psychological interventions on physical exercise self-regulation following coronary
421
rehabilitation. Int J Behav Med 2005; 12:244-55.
422 423 424 425 426 427
AC C
420
29. Robertson D & Keller C. Relationships among health beliefs, self-efficacy, and exercise adherence in patients with coronary artery disease. Heart Lung 1992; 21:58-63. 30. Carroll DL. The importance of self-efficacy expectations in elderly patients recovering from coronary artery bypass surgery. Heart Lung 1995; 24:50-59. 31. O'Farrell P, Murray J, Hotz SB. Psychologic distress among spouses of patients undergoing cardiac rehabilitation. Heart Lung 2000; 29: 97-104.
16
ACCEPTED MANUSCRIPT
428
32. Rohrbaugh MJ, Shoham V, Coyne JC, Cranford JA, Sonnega JS, Nicklas JM. Beyond the
429
‘self’ in self-efficacy: Spouse confidence predicts patient survival following heart failure. J
430
Fam Psychol 2004;18:184-193.
RI PT
431
AC C
EP
TE D
M AN U
SC
432
17
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
433 Table 1. Demographics of Patients and Partners by Study Group 434 435 P Only Group P+P Group Comparison 436 n = 151 n = 150 Test 437 2 Characteristics Frequency (%) Frequency (%) Χ , P-value 438 2 PT Gender 439 χ = 1.98, Male 106 (70.2) 116 (77.3) p = .16 440 Female 45 (29.8) 34 (22.7) 441 2 PR Gender χ = 2.39, 442 Male 45 (29.8) 33 (22.0) p = .12 443 Female 106 (70.2) 117 (78.0) 444 2 PT ICD Indication χ = .03, 445 Primary 91 (60.3) 89 (59.3) p = .87 446 CRT-D 49 (16.5) 42 (14.1) 447 Secondary 60 (39.7) 61 (40.7) 2 448 PT Race/Ethnicity χ = 4.35, 449 138 (92.0) p = .36 Caucasian 136 (90.1) 450 2 (1.3) American Indian/Alaskan 4 (2.6) 0 (0) Asian/Pacific Islander 3 (2.0) 451 Black/African 6 (4.0) 6 (4.0) 452 2 (1.3) 4 (2.7) Hispanic/Latino 453 2 PR Race/Ethnicity χ = 7.17, 454 Caucasian 133 (88.1) 133 (88.7) p = .13 455 American Indian/Alaskan 3 (2.0) 6 (4.0) 456 Asian/Pacific Islander 6 (4.0) 1 (0.7) 457 Black/African 3 (2.0) 7 (4.7) 458 Hispanic/Latino 6 (4.0) 3 (2.0) 2 459 Annual Income χ = 2.98, 460 6 (4.0) p = .70 <$10,000 2 (1.3) 461 $10,000 - $49,999 66 (43.7) 58 (38.7) $50,000 - $89,999 48 (31.4) 45 (30.0) 462 $90,000 or > 32 (21.2) 37 (24.7) 463 Not willing to answer 3 (2.0) 4 (2.7) 464 T-test, 465 mean ± sd mean ± sd P-value 466 PT Age 65.01 ± 11.26 63.26 ± 12.50 t = 1.28, 467 p = .20 468 PR Age 63.46 ± 11.65 61.46 ± 13.06 t = 1.40, 469 p = .16 470 PT Charlson 2.26 ± 1.44 2.32 ± 1.53 t = -0.36, 471 p = .72 472 PR Charlson 0.73 ± 0.99 0.71 ± 1.16 t = 0.12, 473 p = .90 474 PT EF (%) 33.79 ± 13.49 34.38 ± 15.17 t = 0.36, 475 p = .72 476 477 PT=patient, PR=partner, BMI=body mass index, EF=ejection fraction, X2=Chi square. 478
18
ACCEPTED MANUSCRIPT
479
Figure Legends
480 481
Figure 1: Consort Diagram
RI PT
482 Figure 2. Significant changes by study group for patients (Panel A & B) and partners (Panel C &
484
D) across 12 months post-ICD implant.
AC C
EP
TE D
M AN U
SC
483
19
ACCEPTED MANUSCRIPT Patient’s Assessed for Eligibility n = 1015 Eligible n = 861 Not in Study
Not Eligible
n = 560
n = 154
Randomized n = 301
1 Month
1 Month
Completed n=144
Completed n=142
Missing Data Dropped Lost to F/U
5 1 1
3 Month
Completed n=148 Missing Data Lost to F/U
6 Month
6 Month
Completed n=144
TE D
3 1 2 3
Missing Data Dropped Deceased Lost to F/U
12 Month
12 Month
Completed n=138
Completed n=143
EP
Missing Data Dropped Deceased Lost to F/U
1 1
M AN U
5 1 2
Completed n=142 Missing Data Dropped Deceased Lost to F/U
AC C
7 1
3 Month
Fail to complete BL data
Figure 1: Consort Diagram
Missing Data Lost to F/U
Completed n=143 Missing Data Dropped Lost to F/U
Unable to contact
Excluded
RI PT
No Interest Too busy Study protocol too involved Poor Health Other Refuses to sign consent No reason MD does not recommend Partner refused
PT + PR Baseline n=150
SC
Refused
PT Only Baseline n=151
2 3 5 3
Dropped Deceased Lost to F/U
3 1 1 1
2 4 1
Ineligible due to inclusion criteria Not first ICD Did not receive ICD Time since implant too long No English No significant other No phone Ineligible due to exclusion criteria Patient Comorbid Illness Partner Comorbid Illness
ACCEPTED MANUSCRIPT
Patient Symptoms
Patient Depression 6
10
4
9
3
8.5
2
8
1
7.5 Baseline
3 months
0
6 months 12 months
Baseline P+P
Partner Self-Efficacy 9.5
12 months
Panel B. Patient Depression over 12 months by Group, p=0.006
Partner Caregiver Burden
M AN U
Panel A. Patient Symptoms over 12 months by Group, p=0.02
3 months 6 months P Only P+P
SC
P Only
RI PT
5
9.5
2
9
1.5
8.5
1
8
0.5
7.5
0
3 months P Only
6 months 12 months P+P
TE D
Baseline
Panel C. Partner Self-Efficacy over 12 months by Group, p=0.001
AC C
EP
Figure 2.
Baseline
3 months P Only
6 months
12 months
P+P
Panel D. Partner Caregiver Burden over 12 months by Group, p=0.002
S1547-5271(18)31027-0
DOI:
10.1016/j.hrthm.2018.10.011
Reference:
HRTHM 7772
To appear in:
Heart Rhythm
Received Date: 22 February 2018
Please cite this article as: Dougherty CM, Thompson EA, Kudenchuk PJ, Patient Plus Partner Trial: A Randomized Control Trial of Two Interventions to Improve Outcomes following an Initial Implantable Cardioverter Defibrillator, Heart Rhythm (2018), doi: https://doi.org/10.1016/j.hrthm.2018.10.011. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
1 2
Patient Plus Partner Trial: A Randomized Control Trial of Two Interventions to Improve Outcomes following an Initial Implantable Cardioverter Defibrillator
3 Short title: P+P After an ICD
RI PT
4 5
Cynthia M. Dougherty ARNP, PhD, FAHA, FAAN Professor, University of Washington School of Nursing
SC
Elaine A. Thompson RN, PhD Professor Emeritus, University of Washington School of Nursing
Peter J. Kudenchuk, MD, FHRS Professor, University of Washington School of Medicine, Division of Cardiology
M AN U
6 7 8 9 10 11 12 13 14 15 16 17 18
TE D
19 Corresponding Author:
21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37
Cynthia M. Dougherty University of Washington School of Nursing 1959 NE Pacific Street, HSB T615A Box 357266 Seattle, WA 98195 [email protected] (206) 221-7927
AC C
EP
20
Word count: 4927
Funding: NIH, NHLBI R01HL086580-01A2 Clinicaltrials.gov: NCT01252615
The authors report no conflicts of interest.
1
ACCEPTED MANUSCRIPT
Abstract
M AN U
SC
RI PT
Background: Interventions to improve physical and psychological outcomes in recipients with an initial implantable cardioverter (ICD) and their intimate partners are largely unstudied, though likely to have a major impact on adjustment to the ICD and general well-being. Objective: To report the primary outcomes of the Patient Plus Partner randomized control trial. Methods: In a two-group (N=301) prospective RCT, we compared two social-cognitive-based intervention programs [Patient Plus Partner (P+P) and Patient Only (P only)], implemented following initial ICD implant. The patient intervention, consisting of educational materials, nursedelivered telephone coaching, videotape demonstrations, and access to a nurse via a 24/7 pager, was implemented in both groups. P+P also incorporated partner participation. The primary patient outcomes were symptoms and anxiety at 3 months. Other outcomes were physical function (SF-36, ICD shocks-patient); psychological adjustment (PHQ-9); relationship impact (DAS, OCBS-partner); self-efficacy and knowledge (SCA-SE, SCA-OE, KSA); and health care utilization (outpatient visits, hospitalizations), at hospital discharge, 1, 3, 6, and 12 months post-ICD. Results: For patients, P+P compared to P Only was more effective in improving symptoms (P=0.02), depression (P=0.006), self-efficacy (P=.02), outcome expectations (P=0.03), and knowledge (P=0.07). For partners, P+P was more effective in improving partner caregiver burden (P=0.002), self-efficacy (P=0.001) and ICD knowledge (P=0.04). Conclusion: An intervention that integrated the partner into the patient’s recovery following an ICD improved outcomes for both. Beyond survival benefits of the ICD, intervention programs designed to address both the patient and their partner living successfully with an ICD are needed and promising.
EP
TE D
Key Words: Sudden cardiac arrest, ICD, clinical trial, self-efficacy, telephone intervention, partner, anxiety, depression, health care use
AC C
38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69
2
ACCEPTED MANUSCRIPT
INTRODUCTION
70 71
Research points to the important reciprocal influences that patients and partners have on one another’s recovery experiences after a cardiac event [1]. However, interventions
73
designed to impact outcomes of both patients with an implantable cardioverter defibrillator (ICD)
74
and their intimate partners, have not been systematically tested. Given the family stress caused
75
by life-threatening illness and the ever-increasing number of ICDs being implanted worldwide
76
[2], interventions are needed that are timely, both patient and partner-relevant, and integrated in
77
a manner that improves patient recovery. In our past research, we demonstrated that intimate
78
partners of recent ICD recipients reported better physical health than the patient, but their health
79
declined during the first year post-ICD. Intimate partners also reported significantly higher levels
80
of anxiety and depression over the first 12-months, higher than patients themselves. Following
81
ICD implantation, partners consistently reported high caregiving demands and a corresponding
82
lower level of satisfaction with their relationship with the patient [3,4]. Partners also note that
83
important information and support was typically not available to them in outpatient settings.
SC
M AN U
TE D
84
RI PT
72
We previously reported outcomes of a patient-focused self-efficacy intervention following an initial ICD implant after successful resuscitation from sudden cardiac arrest [5,6]. At the
86
conclusion of the intervention, we interviewed patients to determine essential elements of the
87
intervention. Most study participants wanted more telephone contact with the nurse, and would
88
have liked an accompanying videotape to demonstrate important skills; many also desired more
89
psychological support, and requested an intervention to better support their intimate partners.
90
We incorporated these suggestions into our original patient only (P Only) intervention, and
91
created an expanded Patient Plus Partner (P+P) intervention [6]. We then designed and
92
conducted a prospective randomized clinical trial comparing outcomes between the two
93
interventions, one that focused strictly on the ICD recipient (P Only) and one that was expanded
94
to include the partner (P+P). We hypothesized that the expanded (P+P) intervention would
95
result in better physical and psychological outcomes for both patients and partners, compared to
AC C
EP
85
3
ACCEPTED MANUSCRIPT
96
the intervention that focused on the patient alone (P Only). Secondarily, we tested for
97
differences in outcomes by ICD indication (e.g., primary vs. secondary prevention). METHODS
98 Design
RI PT
99
ICD recipients were randomized using a prospective 2-group blocked clinical trial (RCT)
101
design. The strata were Charlson Co-morbidity score (<2 or >2) [10] and ICD indication (primary
102
vs. secondary prevention). The RCT compared two intervention groups: 1) P Only, and 2) P+P
103
during the first year post-ICD implant. Study outcomes were: a) physical functioning in the
104
patient with an ICD (patient concerns/symptoms, general health, ICD shocks) and in the partner
105
(partner concerns/symptoms, general health); b) psychological adjustment (anxiety and
106
depression) in both the patient and partner; c) relationship impact for the patient (dyadic
107
adjustment) and partner (dyadic adjustment, caregiver burden); d) self-efficacy for both the
108
patient and partner, and e) health care utilization (outpatient visits, hospitalizations) for both
109
patient and partner.
110
Sample and Setting
TE D
M AN U
SC
100
Study participants (N=301) were identified as first-time ICD recipients based on
112
systematic screening at 12 acute care institutions in Washington, Oregon, Georgia, and North
113
Carolina between 2010-2015. Identified patients and their partners were approached for study
114
participation while hospitalized for the initial ICD implant. Written consent from both were
115
required for study eligibility. Study procedures were approved by hospital-based IRBs and the
116
academic IRB providing study oversight. Inclusion criteria were: 1) first ICD implant for primary
117
or secondary prevention of sudden cardiac arrest (SCA); 2) ability to read, speak, and write
118
English; 3) access to telephone for 1 year after ICD implant; and 4) having an intimate partner
119
(spouse, lover, life partner or significant other) involved in a committed relationship with the
120
patient. Exclusion criteria were: 1) clinical co-morbidities that impaired cognitive or physical
121
function; 2) short BLESSED cognitive screening score > 6 [7]; 3) age < 21 years; 4) AUDIT-C
AC C
EP
111
4
ACCEPTED MANUSCRIPT
122
score > 4 for alcohol use [8]; and/or 5) ASSIST 2.0 score > 4 for daily non-medical use of illicit
123
substances [9].
124
Protocol and Interventions After baseline data were collected, subjects were randomized to one of the two
126
interventions (P Only or P+P). A computerized program generated a balanced randomization
127
taking into account ICD indication, age, gender, and Charlson co-morbidity index [10]. All
128
participants completed a battery of measurements at the same time periods, which included
129
baseline (hospital discharge), and 1, 3, 6, and 12 month follow-ups, after which they received
130
study-related intervention materials. All participants received usual care from their providers as
131
well as interim telephone calls from study staff every 3 months to collect health care utilization
132
information. A description of the study interventions and intervention fidelity is found in the
133
supplementary materials.
134
Outcome Measures
M AN U
SC
RI PT
125
Outcomes were assessed at baseline (hospital discharge), and at 1 (midway through the
136
intervention, 3 (intervention completion), 6, and 12 months. Outcome measures, summarized in
137
Table S1 (supplemental materials), were obtained for all patients and their respective partners
138
in both intervention groups.
139
Analysis
EP
Prior to the study, we estimated that a total of 300 patients would be needed to provide
AC C
140
TE D
135
141
80% power to detect group differences in the primary outcomes of physical functioning (PCA-
142
symptoms) and psychological adjustment (STAI-anxiety) at 3 months, adjusted for an
143
anticipated 15% dropout. All analyses were conducted using an intent-to-treat approach.
144
Initially, we reviewed data for their distributional properties (means, variances, kurtosis) and
145
potential outliers. Using chi square or t-tests, we tested for group equivalence in baseline
146
demographic and outcome measures. Group intervention effects were tested for change in the
147
primary outcomes from baseline to 3, 6 and 12 months using repeated analysis of co-variance
5
ACCEPTED MANUSCRIPT
(ANCOVA, SPSS 21.0, SPSS Inc, Chicago, IL), controlling for baseline outcome values, age,
149
sex, Charlson score, ejection fraction %, and ICD indication (primary vs. secondary) [24].
150
ANCOVA F statistic indicated if the means between the two groups differed significantly. In
151
conducting secondary comparisons, Bonferroni corrections were made to adjust for potential
152
Type I error by dividing the α value by the number of additional comparisons (α/N) made in each
153
outcome category by timeframe. Statistical significance was defined at α ≤ 0.02, 2-tailed.
154
Intervention effects were also examined by indication for ICD implantation (primary vs.
155
secondary prevention). RESULTS
157
M AN U
156
SC
RI PT
148
Participants
Over the five year study, 1051 patient-partner pairs were approached for study
159
participation, of these 861 were eligible (Figure 1). The patient who received an ICD was the
160
“index case” for study referral. Of the 861 eligible patients, 301 patient-partner pairs met
161
eligibility criteria and were randomized to one of the two study groups: 151 pairs to the P Only,
162
and 150 to the P+P intervention. Of these, 92% of patients in the P Only 95% of patients in the
163
P+P group completed 12-month assessments. The follow-up rate over 12 months for partners
164
was similar to that of patients in both study groups (91% of partners in P Only; 94% of partners
165
in P+P).
167
EP
Patients
AC C
166
TE D
158
Study patients in both P Only and P+P groups averaged 64.14+11.9 (range 26-93)
168
years, had a Charlson co-morbidity score of 2.29 ± 1.49 and a body mass index (BMI) of 29.57
169
± 6.2. The majority were male (73.8%), Caucasian (91.0%), had some college education
170
(50.8%), and were retired from work (43.2%) or disabled (15.6%), and received an ICD for
171
primary prevention of cardiac arrest (58.9%) or secondary prevention of cardiac arrest (40.2%).
172
At baseline enrollment, 37.5% of patients reported a sedentary lifestyle, 20.6% were taking anti-
6
ACCEPTED MANUSCRIPT
173
depressant medications, 11.0% were taking anti-anxiety medications, and 13% were current
174
smokers (Table 1 and Table S2 in supplemental materials). Partners
176
Study partners in both P Only and P+P groups averaged 62.46+12.4 (range 28-91)
RI PT
175
years, had a Charlson co-morbidity score of 0.72+1.08 and a BMI of 28.9 ± 6.5. The majority
178
were female (74.1%), Caucasian (88%), had some college education (39.9%), and were retired
179
from work (40.5%) or disabled (6%). At baseline enrollment, 31.3% of partners reported a
180
sedentary lifestyle, 25.6% were taking anti-depressant medications, 18.9% were taking anti-
181
anxiety medications, and 12.3% were current smokers. (Table 1, and Table S2 supplemental
182
materials).
M AN U
SC
177
By comparison, patients and their partners in P Only and P+P groups did not differ on
183
many baseline factors. However, patients were less likely to be actively employed and their
185
partners more likely to be taking antidepressant medications in the P Only than P+P group;
186
whereas ASSIST recreational drug use was higher among patients and partners in the P+P
187
group.
188
Intervention Fidelity
Over the 5 year course of the trial, the proportion of intervention elements delivered
EP
189
TE D
184
between treatment groups (98.5 +1.2% in P Only and 97.5+2.3% in P+P), and the number of
191
phone calls completed (98.5%+1.2 in P Only and 97.5%+2.3 in P+P) were similar. The
192
cumulative time spent on telephone calls was significantly higher for those in the P+P group
193
(mean+SD) (322+92 minutes vs 250+80 minutes, p=0.001), given that engaging both the
194
partner and patient required more time to deliver the intervention. Partners who were
195
randomized to P+P participated in an additional 291+77 cumulative minutes in the PTG
196
intervention; 85% of whom completed all of the planned telephone calls. Most patients (97%)
197
and partners (96%) were favorably impressed by the program and recommended that others
198
participate in a similar one, if offered. Variability in intervention delivery by nurses within the
AC C
190
7
ACCEPTED MANUSCRIPT
intervention groups did not differ significantly, and the effect of the nurse on patient outcomes
200
was not significant.
201
Outcomes
202
Physical Function
203
RI PT
199
Patient Physical Function. The frequency of self-reported PCA symptoms (primary outcome) fell significantly between the intervention groups (Figure 2A and Table S3,
205
supplemental materials) at 12 months, with patients in P+P reporting fewer symptoms than
206
those in P Only (P = 0.02). Fear of dying and distress reduced in both groups over time.
207
Patients in both intervention groups reported lower levels of physical (PCS) than mental health
208
(MCS) at study enrollment and throughout 12 months, with no statistically significant by group
209
differences noted over 12 months. The number of ICD shocks by intervention group did not
210
differ across time, ranging from 3 to 9% over the course of 12 months. Overall ICD shock rates
211
were low, perhaps related to changes in ICD programming standards that occurred during the
212
course of the study.
M AN U
TE D
213
SC
204
Partner Physical Function. As well, partners in both intervention groups reported lower levels of physical (PCS) vs. mental (MCS) health throughout the study, with no statistically
215
significant by group differences noted. The number of physical symptoms reported by partners
216
was lower than patients, and overall general health of partners was better than patients, from
217
enrollment to final follow-up (Table S4, supplemental materials). The number of self-reported
218
symptoms among partners did not differ significantly between the intervention groups over 12
219
months. ICD and health-related partner concerns declined in both intervention groups over time.
220
Psychological Adjustment
221
AC C
EP
214
Patient anxiety (STAI) declined over time (primary outcome), with no significant
222
differences observed between intervention groups (Table S3). However, for P+P vs. P Only,
223
patient depression and depression severity (Figure 2B), significant reductions (P = 0.006 and
224
0.01, respectively) were noted at 12 months. Conversely, partner anxiety and depression
8
ACCEPTED MANUSCRIPT
225
remained relatively stable across the 12-months, without significant differences between P+P
226
and P Only.
227
Relationship Impact Patient Relationship Impact. No statistically significant differences were observed over
RI PT
228
time for P+P vs. P Only with respect to overall dyadic adjustment or subscales of dyadic
230
adjustment, including satisfaction, cohesion, consensus, or affectionate expression among
231
study patients (Table S3). Nor were there significant group differences in patient sexual
232
concerns, which were highest at hospital discharge and declined only modestly over time.
SC
229
Partner Relationship Impact. Similar to patients, there were no statistically significant
234
intervention group differences among partners in overall or subscale measurement of dyadic
235
adjustment, or sexual concerns (Table S4). Partner sexual concerns were higher than patients
236
during the first 3 months post-ICD, and reduced to levels similar to those of patients at 6
237
months. For caregiving burden, however, partners in P+P compared to P Only reported
238
significantly lower levels of caregiver burden and time spent in caregiving (Figure 2D). There
239
were also trends toward reduction in caregiving difficulty reported by partners in the P+P
240
compared to P Only at 3, 6, and 12 months, (P = 0.06, 0.07, and 0.06, respectively).
241
Self-Efficacy
TE D
EP
242
M AN U
233
Patient Self-Efficacy. Self-efficacy was assessed using three indicators: self-efficacy for living with the ICD, outcome expectations after an ICD, and ICD knowledge. For specific
244
aspects of self-efficacy for living with an ICD measured at 12 months, for P+P vs. P Only there
245
were significantly higher levels in ICD function (P = 0.01), managing physical changes caused
246
by the ICD (F=10.86, P = 0.001), relationship management related to the ICD (F=4.66, P =
247
0.03), and a trend for ICD safety (F=2.90, P= 0.09). For outcome expectations, a trend toward
248
improvement for patients in P+P vs. P Only was observed at 3 months (F=3.34, P = 0.07),
249
which was more pronounced at 6 months (F=5.06, P = 0.03), but declined by 12 months
AC C
243
9
ACCEPTED MANUSCRIPT
250
(F=0.36, P = 0.55). Patient ICD knowledge did not significantly differ over time for P+P vs. P
251
Only, but was slightly higher in P+P at most points in time. Partner Self-Efficacy. For partner self-efficacy for living with the ICD, the P+P partner
253
group reported significantly higher self-efficacy at all data collection points across 12 months
254
(Figure 2C; Table S4). For outcome expectations, there were no significant group differences
255
across time. Partner ICD knowledge was significantly higher in P+P vs. P Only at all time points
256
over 12 months.
257
Health Care Utilization
SC
Patient Health Care Utilization. No statistically significant group differences in the
M AN U
258
RI PT
252
frequency of patient hospitalization were noted over 1 year (Table S3). The patient
260
hospitalization rate in both groups was 7.9% in the first month post-ICD, 20.6% at 1-3 months,
261
23.3% at 3-6 months, and 36.9% at 6-12 months. The average number of patient hospitalization
262
days did not differ significantly by treatment group over 12 months, with total hospital days in
263
both groups ranging from 2-6 days. Patients were hospitalized during the first year post-ICD for
264
cardiac issues such as chest pain, heart failure exacerbation, cardiac arrhythmias, or ICD
265
related reasons. However, patients in P+P had fewer outpatient visits over the course of the trial
266
than those in P Only (2.54 vs. 3.09, P = 0.03; data not shown), which were most often ICD-
267
related for both groups.
EP
Partner Health Care Utilization. No statistically significant differences in the frequency of
AC C
268
TE D
259
269
partner hospitalizations were found between treatment groups during the follow-up year (Table
270
S4). The hospitalization rate among partners in both groups was 2.7% during the first month,
271
3.3% from 1-3 months, 11.3% from 3-6 months, and 19.3% from 6-12 months, primarily for non-
272
cardiac surgical procedures. The average number of hospitalization days in partners did not
273
differ significantly by group over the 12 months, ranging from 2-5 days. No significant group
274
differences were observed in the number of outpatient visits by partners over 12 months. On
10
ACCEPTED MANUSCRIPT
275
average, partners in both groups had 1 outpatient visit in each 3-month interval across the 12
276
months.
277
ICD Indication: Primary vs. Secondary Prevention We conducted exploratory analyses to determine the effects of ICD indication (primary
RI PT
278
vs. secondary prevention) on differential benefit from the interventions (data not shown). There
280
were marked baseline differences in the health of patients receiving an ICD for primary vs.
281
secondary prevention reflected in higher Charlson co-morbidity scores (t = 2.88, P = 0.004),
282
higher BMI (t = 2.17, P = 0.03), and lower ejection fraction (t = 8.64, P = 0.001) in the primary
283
prevention group. Outcomes at 3 months showed that patients receiving an ICD for secondary
284
prevention, had greater reductions in symptoms (PCA, F = 7.70, P = 0.006), better physical
285
health (SF-36 PCS, F = 8.91, P = 0.001), and greater ICD knowledge (F = 5.02, P = 0.02).
286
These changes were sustained at 12 months. Patients who received an ICD for primary
287
prevention experienced more hospitalizations (F = 5.48, P = 0.04) over the first 3 months, but
288
this difference was no longer apparent by 12 months. By comparison, ICD indication did not
289
appear to have a robust or consistent influence on partner responses over the course of the
290
trial.
292
M AN U
TE D
DISCUSSION
EP
291
SC
279
The health impact of an ICD reaches far beyond rhythm management alone, extending to the functional and psychological aspects of daily living for both patients and their intimate
294
partners. The importance of this RCT lies in being the first to establish such benefits from an
295
intervention directed at both the patient and partner. We demonstrated that a P+P intervention
296
is superior to a P Only intervention in reducing patient symptoms and depression, while
297
improving self-confidence and ICD knowledge over time. The P+P intervention also proved to
298
be superior to a P Only approach for the partner in reducing caregiver burden, while
299
concurrently improving self-confidence and knowledge.
AC C
293
11
ACCEPTED MANUSCRIPT
300
The intervention framework was based on Social Cognitive Theory [11], which is well suited for patients undergoing initial ICD implant and their partners. The ICD implant recovery
302
period requires acquisition of new knowledge, behavioral skills and confidence to promote living
303
successfully with an implanted device that is known to impact quality of life. A central
304
component of this framework is self-efficacy, the belief in one’s capacity to engage in new
305
behaviors in new situations. Enhancing self-efficacy as a means to facilitate health behavior
306
change in CV disease has been studied extensively, demonstrating that self-efficacy influences
307
cardiac-related lifestyle changes [25], motivation, and emotional reactions to stressful situations
308
[26]. For patients recovering from acute myocardial infarction (MI) and heart surgery, self-
309
efficacy have been positively associated with adherence to exercise programs [27-30].
310
Specifically, with coaching from health care providers, self-efficacy for walking and running
311
generalized to activities such as weight lifting and sexual intercourse [31]. Higher self-efficacy in
312
partners has been shown to predict long-term survival in patients with heart failure [32].
SC
M AN U
The content of the present intervention was tested earlier by our research group in
TE D
313
RI PT
301
patients undergoing initial ICD implant for secondary prevention [5,6]. The current study
315
included patients who had received an ICD for secondary as well as primary prevention.
316
Comparing study outcomes with our previous work [5,6], we note the following similarities in
317
both studies: 1) patients report significant reductions in symptoms and improved self-efficacy
318
and knowledge over 12 months, and 2) patient outcomes improved the most between hospital
319
discharge and 3 months, and then stabilized over the remainder of the year.
AC C
320
EP
314
Analyses based on primary vs. secondary indication demonstrated that reasons for
321
receiving an ICD were associated with differential response to the intervention in patients. The
322
patient group receiving an ICD for secondary prevention demonstrated a stronger response to
323
the intervention, whereas there were few differences in partner responses to the intervention
324
based on ICD indication.
325
Strengths and Limitations
12
ACCEPTED MANUSCRIPT
The strengths of this study were its randomized design, careful and detailed attention to
327
intervention fidelity, and low drop-out rates in both patient and partner groups. Few intervention
328
programs for patients who get an initial ICD have been developed and tested, and until now
329
none have included the partner. The interventions were delivered in the outpatient setting using
330
telephone delivery, without the need for costly in person clinic visits. The interventions are being
331
further individualized and developed for implementation into routine clinical care.
Though randomized, this trial was not blinded. Study outcomes were analyzed without
SC
332
RI PT
326
knowledge of the subject’s group assignment. The study design did not include a true patient
334
control group for comparison against a patient intervention group. However, the trial did include
335
a partner usual care group. Some study outcomes relied on self-report. In anticipation of this
336
limitation, individuals were excluded from study participation if they had cognitive deficits, were
337
in extremely poor physical health after ICD implant, or lacked a intimate partner who was willing
338
to participate. Study results pertain principally this population. Finally, while patients in the P
339
Only group were asked not to share intervention content with their partners, we had little control
340
over partner exposure to the written intervention content. However, partner participation in the
341
telephone portion of the P Only intervention was monitored to ensure that partners did not
342
participate in the weekly calls.
EP
TE D
M AN U
333
343
A telephone delivered self-efficacy based intervention that included both the patient and
AC C
344
CONCLUSIONS
345
the intimate partner following implantation of an initial ICD, resulted in significant reductions in
346
patient symptoms and depression, with higher self-efficacy, outcome expectations, and
347
knowledge over the P only intervention. Benefits to the partner included a reduction in caregiver
348
burden and improved self-efficacy and knowledge to care successfully for the patient with an
349
ICD. Post-hospital based intervention programs established for those with an ICD should
350
include both the patient and their intimate partner in education, support, and home care
351
management.
13
ACCEPTED MANUSCRIPT
References
352
354 355
1. Van Horn E, Fleury J, Moore S. Family interventions during the trajectory of recovery from cardiac event: an integrative literature review. Heart Lung 2002; 31:186-98. 2. Munir MB, Alqahtani F, Aljohani S, Bhirud A, Modi S, Alkhouli M. Trends and predictors of
RI PT
353
implantable cardioverter defibrillator implantation after sudden cardiac arrest: Insight from the
357
national inpatient sample. Pacing Clin Electrophysiol. 2018 Jan 9. doi: 10.1111/pace.13274.
358
[Epub ahead of print].
359
SC
356
3. Dougherty CM, Fairbanks AM, Eaton LH, Morrison MK, Kim MS, Thompson EA. Comparison of patient and partner quality of life and health outcomes in the first year after an
361
implantable cardioverter defibrillator (ICD). J Behav Med 2016, 39:94-106.
M AN U
360
362
4. Dougherty CM, Thompson EA. Intimate partner physical and mental health after sudden
363
cardiac arrest and receipt of an implantable cardioverter defibrillator. Res Nursing Health
364
2009; 32:432-442.
366 367
5. Dougherty CM, Thompson EA, Lewis FM. Long term outcomes of a nursing intervention
TE D
365
after an ICD. Pacing Clin Electrophysiol 2005; 28:1157-67. 6. Dougherty CM, Lewis FM, Thompson ET, Baer, JD, Kim W. Short term efficacy of a telephone intervention after ICD implantation. Pacing Clin Electrophysiol, 2004; 27:1594-
369
1602.
7. Blessed G, Tomlinson BE, Roth M. The association between quantitative measures of
AC C
370
EP
368
371
dementia and of senile change in the cerebral grey matter of elderly participants. British J
372
Psychia 1986;512:797-811.
373 374 375 376 377
8. Bush K, Kivlahan DR, McDonell MB, Fihn SD, Bradley KA. The AUDIT Alcohol Consumption Questions (AUDIT-C). Arch Intern Med 1998;158:1789-1795. 9. WHO Assist Working Group. The Alcohol, smoking, and substance involvement screening test (ASSIST): Development, reliability, and feasibility. Addict 2002; 97:1183-94. 10. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic
14
ACCEPTED MANUSCRIPT
378
comorbidity in longitudinal studies: Development and validation. J Chronic Dis 1987;40:373-
379
383.
381 382
11. Bandura A. Self-efficacy: Toward a unifying theory of behavioral change. Psych Rev 1977; 84:191-215.
RI PT
380
12. Dougherty CM, Benoliel JQ, Bellin CM. Domains of nursing intervention after sudden
cardiac arrest and automatic internal cardioverter defibrillator implantation. Heart Lung 2000;
384
29:79-86.
SC
383
13. Dougherty CM, Pyper GP, Benoliel JQ. Domains for nursing intervention for intimate
386
partners of sudden cardiac arrest survivors. J of Cardiovac Nurs 2004; 19:21-31.
387
M AN U
385
14. Dougherty CM, Thompson EA, Kudenchuk PJ. Development and Testing of an Intervention
388
to Improve Outcomes for Partners following Receipt of an Implantable Cardioverter
389
Defibrillator (ICD) in the Patient. Adv Nurs Sci 2012, 35:359-377. 15. Jenkins LS. Quality of life in patients of sudden cardiac death entering treatment. In Dunbar
391
SB, Ellenbogen KA, Epstein AE. Sudden Cardiac Death: Past, Present, and Future. Armonk:
392
NY: Futura Publishing;1997,289-305.
395 396 397 398 399 400 401 402
1987; 40:473-80.
EP
394
16. Ware JE. Standards for validating health measures: Definitions and content. J Chronic Dis
17. Spielberger C, Gorsuch R, Lushene R. Manual for the state-trait anxiety inventory (selfevaluation questionnaire). Palo Alto, CA: Consulting Psychologists, 1970.
AC C
393
TE D
390
18. Kroenke K, Spitzer RL, Williams JB. The PHQ-9: validity of a brief depression severity measure. J Gen Int Med 2001;16:606-613. 19. Steinke EE, Gill-Hopple K, Valdez D, Wooster M. Sexual concerns and educational needs after an implantable cardioverter defibrillator. Heart Lung 2005: 34:299-308. 20. Steinke EE. Sexual concerns of patients and partners after an implantable cardioverter defibrillator. Dimens Crit Care Nurs 2003; 22:89-96.
15
ACCEPTED MANUSCRIPT
403 404 405
21. Spanier G.B. Assessing dyadic adjustment: New Scales for assessing the quality of marriage and similar dyads. J Marriage Fam 1976; 38:15-28. 22. Bakas T, Austin JK, Jessup SL, Williams LS, Oberst MT. Time and difficulty of tasks provided by family caregivers of stroke survivors. J Neurosci Nurs 2004, 36: 95-105.
407
23. Dougherty CM, Thompson EA, Johnston SA. Reliability and validity of the self-efficacy
RI PT
406
408
expectations and outcome expectations after ICD implantation scales. Appl Nurs Res 2007;
409
20:116-124.
Saddle River, NJ: Pearson/ Prentice Hall, 2004.
SC
411
24. Kepple G, Wickens TD. Design and analysis: A researcher’s handbook (4th Ed). Upper
M AN U
410
412
25. Bandura A. Self-efficacy mechanism in human agency. Am Psych 1982; 37:122-147.
413
26. Izawa KP, Watanabe S, Omiya K, Hirano Y, Oka K, Osada N, Iijima S. Effect of the self-
414
monitoring approach on exercise maintenance during cardiac rehabilitation: A randomized
415
controlled trial. Am J Phys Med Rehab 2005; 84:313-21.
27. Taylor CB, Bandura A, Ewart CK, Miller NH, DeBusk RF. Exercise testing to enhance wives
TE D
416 417
confidence in their husbands cardiac capability soon after clinically uncomplicated acute
418
myocardial infarction. Am J Card 1985; 55:635-38. 28. Sniehotta FF, Scholz U, Schwarzer R, Fuhrmann B, Kiwus U, Voller H. Long term effects of
EP
419
two psychological interventions on physical exercise self-regulation following coronary
421
rehabilitation. Int J Behav Med 2005; 12:244-55.
422 423 424 425 426 427
AC C
420
29. Robertson D & Keller C. Relationships among health beliefs, self-efficacy, and exercise adherence in patients with coronary artery disease. Heart Lung 1992; 21:58-63. 30. Carroll DL. The importance of self-efficacy expectations in elderly patients recovering from coronary artery bypass surgery. Heart Lung 1995; 24:50-59. 31. O'Farrell P, Murray J, Hotz SB. Psychologic distress among spouses of patients undergoing cardiac rehabilitation. Heart Lung 2000; 29: 97-104.
16
ACCEPTED MANUSCRIPT
428
32. Rohrbaugh MJ, Shoham V, Coyne JC, Cranford JA, Sonnega JS, Nicklas JM. Beyond the
429
‘self’ in self-efficacy: Spouse confidence predicts patient survival following heart failure. J
430
Fam Psychol 2004;18:184-193.
RI PT
431
AC C
EP
TE D
M AN U
SC
432
17
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
433 Table 1. Demographics of Patients and Partners by Study Group 434 435 P Only Group P+P Group Comparison 436 n = 151 n = 150 Test 437 2 Characteristics Frequency (%) Frequency (%) Χ , P-value 438 2 PT Gender 439 χ = 1.98, Male 106 (70.2) 116 (77.3) p = .16 440 Female 45 (29.8) 34 (22.7) 441 2 PR Gender χ = 2.39, 442 Male 45 (29.8) 33 (22.0) p = .12 443 Female 106 (70.2) 117 (78.0) 444 2 PT ICD Indication χ = .03, 445 Primary 91 (60.3) 89 (59.3) p = .87 446 CRT-D 49 (16.5) 42 (14.1) 447 Secondary 60 (39.7) 61 (40.7) 2 448 PT Race/Ethnicity χ = 4.35, 449 138 (92.0) p = .36 Caucasian 136 (90.1) 450 2 (1.3) American Indian/Alaskan 4 (2.6) 0 (0) Asian/Pacific Islander 3 (2.0) 451 Black/African 6 (4.0) 6 (4.0) 452 2 (1.3) 4 (2.7) Hispanic/Latino 453 2 PR Race/Ethnicity χ = 7.17, 454 Caucasian 133 (88.1) 133 (88.7) p = .13 455 American Indian/Alaskan 3 (2.0) 6 (4.0) 456 Asian/Pacific Islander 6 (4.0) 1 (0.7) 457 Black/African 3 (2.0) 7 (4.7) 458 Hispanic/Latino 6 (4.0) 3 (2.0) 2 459 Annual Income χ = 2.98, 460 6 (4.0) p = .70 <$10,000 2 (1.3) 461 $10,000 - $49,999 66 (43.7) 58 (38.7) $50,000 - $89,999 48 (31.4) 45 (30.0) 462 $90,000 or > 32 (21.2) 37 (24.7) 463 Not willing to answer 3 (2.0) 4 (2.7) 464 T-test, 465 mean ± sd mean ± sd P-value 466 PT Age 65.01 ± 11.26 63.26 ± 12.50 t = 1.28, 467 p = .20 468 PR Age 63.46 ± 11.65 61.46 ± 13.06 t = 1.40, 469 p = .16 470 PT Charlson 2.26 ± 1.44 2.32 ± 1.53 t = -0.36, 471 p = .72 472 PR Charlson 0.73 ± 0.99 0.71 ± 1.16 t = 0.12, 473 p = .90 474 PT EF (%) 33.79 ± 13.49 34.38 ± 15.17 t = 0.36, 475 p = .72 476 477 PT=patient, PR=partner, BMI=body mass index, EF=ejection fraction, X2=Chi square. 478
18
ACCEPTED MANUSCRIPT
479
Figure Legends
480 481
Figure 1: Consort Diagram
RI PT
482 Figure 2. Significant changes by study group for patients (Panel A & B) and partners (Panel C &
484
D) across 12 months post-ICD implant.
AC C
EP
TE D
M AN U
SC
483
19
ACCEPTED MANUSCRIPT Patient’s Assessed for Eligibility n = 1015 Eligible n = 861 Not in Study
Not Eligible
n = 560
n = 154
Randomized n = 301
1 Month
1 Month
Completed n=144
Completed n=142
Missing Data Dropped Lost to F/U
5 1 1
3 Month
Completed n=148 Missing Data Lost to F/U
6 Month
6 Month
Completed n=144
TE D
3 1 2 3
Missing Data Dropped Deceased Lost to F/U
12 Month
12 Month
Completed n=138
Completed n=143
EP
Missing Data Dropped Deceased Lost to F/U
1 1
M AN U
5 1 2
Completed n=142 Missing Data Dropped Deceased Lost to F/U
AC C
7 1
3 Month
Fail to complete BL data
Figure 1: Consort Diagram
Missing Data Lost to F/U
Completed n=143 Missing Data Dropped Lost to F/U
Unable to contact
Excluded
RI PT
No Interest Too busy Study protocol too involved Poor Health Other Refuses to sign consent No reason MD does not recommend Partner refused
PT + PR Baseline n=150
SC
Refused
PT Only Baseline n=151
2 3 5 3
Dropped Deceased Lost to F/U
3 1 1 1
2 4 1
Ineligible due to inclusion criteria Not first ICD Did not receive ICD Time since implant too long No English No significant other No phone Ineligible due to exclusion criteria Patient Comorbid Illness Partner Comorbid Illness
ACCEPTED MANUSCRIPT
Patient Symptoms
Patient Depression 6
10
4
9
3
8.5
2
8
1
7.5 Baseline
3 months
0
6 months 12 months
Baseline P+P
Partner Self-Efficacy 9.5
12 months
Panel B. Patient Depression over 12 months by Group, p=0.006
Partner Caregiver Burden
M AN U
Panel A. Patient Symptoms over 12 months by Group, p=0.02
3 months 6 months P Only P+P
SC
P Only
RI PT
5
9.5
2
9
1.5
8.5
1
8
0.5
7.5
0
3 months P Only
6 months 12 months P+P
TE D
Baseline
Panel C. Partner Self-Efficacy over 12 months by Group, p=0.001
AC C
EP
Figure 2.
Baseline
3 months P Only
6 months
12 months
P+P
Panel D. Partner Caregiver Burden over 12 months by Group, p=0.002