Impact of Cardiac Devices on the Quality of Life in Pediatric PatientsClinical Perspective
Background—Cardiac rhythm devices are increasingly used in the pediatric population, although their impact on quality of life (QOL) is poorly understood. The purpose of this study was to compare (QOL) scores among pediatric device patients, healthy controls, and congenital heart disease (CHD) patients and determine the key drivers of QOL in pediatric device patients.
Methods and Results—Multicenter, cross-sectional study at 8 pediatric centers of subjects aged 8 to 18 years with either a pacemaker or defibrillator was carried out. Patient–parent pairs completed the Pediatric Quality of Life Inventory and Pediatric Cardiac Quality of Life Inventory. QOL outcomes in device patients were compared with healthy controls and patients with various forms of CHD. Structural equation modeling was used to test for differences in Pediatric Cardiac Quality of Life Inventory scores among (1) device type, (2) presence of CHD, and (3) hypothesized key drivers of QOL. One hundred seventy-three patient–parent pairs (40 defibrillators/133 pacemakers) were included. Compared with healthy controls, patients with devices and their parents reported significantly lower Pediatric Quality of Life Inventory scoring. Similarly, compared with patients with mild forms of CHD, parents and patients with devices reported significantly lower Pediatric Cardiac Quality of Life Inventory scores and were similar to patients with more severe CHD. Key drivers of patient QOL were presence of implantable cardioverter-defibrillator and CHD. For patients, self-perception was a key driver of lower QOL, whereas for parents behavioral issues were associated with lower QOL.
Conclusions—Patient QOL is significantly affected by the presence of cardiac rhythm devices. Whether these effects can be mitigated through the use of psychotherapy needs to be assessed.
Trends in cardiac rhythm device implantation in pediatric populations have demonstrated increasing utilization.1,2 Despite advancements in both generator and lead technology, long-term consequences of device management continue to affect patient outcome.3–5 As device technology and implant indications evolve, there remains a paucity of data in the literature on quality of life (QOL) in the pediatric device population. QOL studies in adult pacemaker cohorts have focused primarily on device-related QOL improvement in symptomatic cohorts and less on the potential detrimental side effects associated with a permanent device.6 In pediatric patients, there is little known about the psychosocial response to pacemaker implantation and no data on the effect of pacemakers on QOL.7,8 Whether or not pacemakers demonstrate similar positive effects on QOL in pediatric populations compared with that seen in adults remains unknown. In adult patients with implantable cardioverter-defibrillators (ICDs), patients have demonstrated increased anxiety, which was worsened in those patients with a history of device shock.9–11 Two previous pediatric-specific ICD studies demonstrated lower QOL in pediatric patients with ICDs compared with healthy controls validating concerns with regard to this patient population.12,13 Device-related complications, including inappropriate device therapy, need for lead, and generator revision, as well as lifestyle modifications such as activity restrictions and resultant cosmetic changes, continue to present a significant concern for patients and practitioners.4 The effect of these morbidities on pediatric patients’ self-perception and behavioral characteristics remains poorly defined.
Clinical Perspective on p 1072
The purpose of this study was to compare QOL scores among pediatric cardiac rhythm device patients, healthy controls, and congenital heart disease (CHD) patients. In addition, we sought to identify key drivers of QOL in pediatric patients with implanted devices. By identifying areas of potential impairment and the mechanistic underpinnings that lead to worsened QOL, we hope to develop strategies to positively impact the management and improve the lives of this population.
This was a multi-institution retrospective cross-sectional study with patients derived from 8 tertiary-care pediatric cardiology centers and conducted in accordance with the Cincinnati Children’s Hospital Medical Center Institution Review Board (No. 2010–2211). This patient population was derived from the Pediatric Cardiac Quality of Life Inventory (PCQLI) Validation Study data registry.14 Patient–parent pairs 8 to 18 years of age who had undergone implantation of a permanent pacemaker or ICD system before QOL assessment were included in analysis. These QOL, self-perception, and behavioral and emotional functioning instruments are reliable and valid and have been previously used in pediatric cohorts.14,15
Patient–parent pairs completed the Pediatric Quality of Life Inventory (PedsQL) and PCQLI to assess both generic and cardiac-specific patient QOL; Self-Perception Profile for Children and Adolescents to assess patient self-perception; and the Child Behavior Checklist and Youth Self-Report to assess patient behavioral/emotional functioning. The PedsQL generates a Total score and Physical Health Summary and Psychological Health Summary subscale scores. The Psychological Health Summary score is a composite of emotional, social, and school functioning.16 The PCQLI generates a Total score and Disease Impact and Psychosocial Impact subscale scores. Disease Impact Subscale score assesses disease state and physical functioning, and Psychosocial Impact Subscale score assesses psychological status and social functioning.14
Patient and Device System Characteristics
Data collected from the PCQLI Validation Study were included up to the point of patient enrollment in the PCQLI Validation Study when QOL assessment was performed. In addition, other pertinent electrophysiology data were obtained from each of the participating centers. Patients were classified into 2 groups based on device type: pacemaker versus ICD. Data were obtained with reference to device pacing (single chamber, dual chamber, or biventricular) and device lead type (transvenous or epicardial). The presence or absence of CHD was analyzed, and patients with CHD were categorized as either single- or 2-ventricle physiology. Patient comorbidities were analyzed, including sinus node dysfunction, congenital heart block, postoperative heart block, cardioinhibitory syncope, and atrial tachycardia. Time from diagnosis was defined as the time from diagnosis of cardiovascular disease to the time of patient intake.
Device-related complications included complications that were a result of the device implant procedure or developed as a part of long-term management of the device, including infection, endocarditis, diaphragmatic pacing, pocket hematoma, lead fracture, lead dislodgment, and device system recall. Pacemaker dependency was defined as the presence or absence of an underlying escape rhythm that would prevent sudden death as judged by the electrophysiological physician at the time of PCQLI testing. Age at device implantation was defined as the patient age at the time of first device system implantation. Similarly, time since device implantation was defined as the time from implantation of initial device to the time of patient intake analysis.
In the ICD group, the indication for device implantation and history of device therapies were obtained. Indication for ICD implantation was classified as primary versus secondary prevention by the electrophysiological physician. Device therapies were classified as appropriate versus inappropriate, based on review of the device therapy episode.
Analysis was performed by comparing patients with cardiac rhythm devices with a healthy control population (online-only Data Supplement Table I). t tests were used to compare the national population of healthy controls presented in the initial PedsQL psychometric article and the PedsQL data from within our PCQLI Validation Study.16 Using the generic version of the PedsQL, 2 separate analyses were conducted. First, scores for healthy children/adolescents were compared with our sample of children/adolescents. Second, parent-proxy reports of both healthy controls and our sample were compared. Throughout the article, the comparison of the device group with the healthy control population is referred to as comparison with healthy controls.
After comparing patients with cardiac rhythm devices with a healthy control population, the device group was compared with patients with variable forms of CHD without devices using the PCQLI. Because the PCQLI is a disease-specific cardiac QOL measurement tool, no comparison with a healthy, cardiac disease-free, population was possible. Patients with variable forms of CHD were chosen for comparison, including patients with isolated bicuspid aortic valve (BAV), tetralogy of Fallot (TOF), and single-ventricle patients s/p palliation with a Fontan procedure to represent a spectrum of CHD severity. Patients with isolated BAV (no aortic stenosis or insufficiency) were used as the closest approximation to a healthy control population for purposes of PCQLI comparison, and the TOF and Fontan groups were used to depict groups with increasing CHD severity. Two-way ANOVA using Tukey test of additivity was used to test for differences between PCQLI QOL scores in patients with devices compared with patients with various forms of CHD, including isolated BAV, TOF, and single-ventricle patients s/p palliation with a Fontan procedure. Because of multiple comparisons, the P value considered statistically significant was adjusted accordingly using a Bonferroni correction. For the purposes of PCQLI comparison, patient groups were mutually exclusive and patients with TOF or Fontan who had a device were included in their respective pacemaker or ICD device group. CHD comparison groups (BAV, TOF, and Fontan) were obtained from the PCQLI Validation Study and did not include patients with devices. If any device group had a significant proportion of associated CHD (>25%), further analysis was performed comparing patients with and without CHD within a given device group. Pearson correlation was used to evaluate the effect of patient age at the time of device implant and time from initial device implant on patient QOL measures. Within the ICD group, Total PedsQL and Total PCQLI scores were compared in patients with and without a history of ICD shock.
Structural equation modeling was used to identify key drivers of QOL in the cardiac rhythm device population. (Figure 1) For this analysis, only the structural modeling component of structural equation modeling or path model was used.17 The path model was evaluated using measures of global fit, using not only the χ2 statistic to determine model fit but also the goodness-of-fit index ≥0.95,18 adjusted goodness-of-fit index ≥0.90,19 and the root mean square error of approximation ≤0.08.20 Standardized coefficients are reported, and statistical significance was determined using an α level of 0.05. All data were analyzed using SAS 9.2 (including PROC CALIS).
Population Demographic Characteristics
There were 173 patient–parent pairs (40 patients with ICDs, 133 patients with pacemakers) (Table 1). The median age was 13 years (8–18 years), and 50% were men. There were no significant differences between pacemaker and ICD groups for age at enrollment, sex, parental income quartile, or cardiac function. Patients with a pacemaker had younger age at the time of their initial device implantation (5.2 versus 10.7 years; P<0.0001) and had a longer time from initial device implantation (10.8 versus 5.5 years; P<0.0001). There was a higher proportion of patients with underlying conduction system abnormalities, including sinus node dysfunction and heart block in the pacemaker cohort. ICD patients had a greater frequency of single-chamber systems compared with pacemaker patients (P=0.02). In ICD patients, 50% of devices were implanted for primary prevention indications. Compared with both ICD and pacemaker groups, the BAV group was younger (P=0.025) but had a greater time since their initial cardiac diagnosis (P<0.0001).
Predictor and Moderator Variable Description
Fifty percent of patients had CHD, with a greater proportion of pacemaker patients having CHD compared with ICD patients (59% versus 20%; P<0.0001; Table 2). In patients with CHD, 63% had biventricular physiology, with no significant difference in the proportion of single versus biventricular patients between device groups (P=0.25). Pacemaker patients were more likely to have an epicardial system (47.4% versus 12.8%; P<0.0001) compared with ICD patients likely secondary to younger age at initial implantation. Thirty-eight percent of patients had ≥1 device-related complication. There was no significant difference between the pacemaker and ICD groups with respect to total device-related complications (39% versus 35%; P=0.64). In both groups, the majority of complications were related to the mechanical device system, either a recall of part of the system or lead fracture/dislodgment. Pacemaker patients were more likely to have an infection or lead fracture, whereas ICD patients were more likely to have had a device system recall. In pacemaker patients, the majority of lead fractures were associated with epicardial leads (31/38; 82%). Within the ICD group, 43% of patients had received ≥1 device shock. Almost half of the delivered therapies (47%) were classified as inappropriate.
There were no differences between the pacemaker and ICD groups for any self-perception or behavioral measures. Both device groups demonstrated self-perception and behavioral measures similar to previously published measures in healthy patients and with no significant difference between device groups for measures of self-perception and behavioral and emotional functioning (data not shown).21 Compared with the entire group of device patients, the BAV group had less total problems (P=0.04), less somatic complaints on the syndrome scales (P=0.03), less evidence for somatic disorder on the Diagnostic and Statistical Manual–oriented scales (P=0.02), and better school competency (P=0.004).
PedsQL: Generic QOL Measure
Compared with healthy controls, patient- and parent-proxy–reported PedsQL scores were significantly lower in patients with pacemakers and ICDs (Figure 2A and 2B). This difference was seen for Total as well as the Health Summary subscale scores. Because the pacemaker group had a significant proportion of patients with associated CHD (59%), analyses were performed dividing the pacemaker group into 2 groups: (1) pacemaker patients with CHD, and (2) pacemaker patients without CHD (Figure 2C and 2D). Pacemaker patients with and without CHD had significantly lower QOL scores compared with the healthy control population. This difference was seen in patient- and parent-proxy–reported PedsQL Total and subscale scores.
PCQLI: Disease-Specific Cardiac QOL Measure
Disease-specific PCQLI scores also demonstrated significant effects imparted by the presence of cardiac devices on QOL (Figure 3). For patient- and parent-proxy–reported Total and Psychosocial Impact scores, both device groups had lower QOL scores compared with patients with isolated BAV and were similar to patients with TOF. For the patient- and parent-proxy–reported Disease Impact score, both device groups had significantly lower scores compared with patients with BAV and TOF. For patient-reported Total and Disease Impact subscale scores and parent-proxy–reported Disease Impact subscale score, pacemaker patients had higher scores compared with Fontan patients, whereas ICD patients and parent-proxy reporters had scores that were similar to the Fontan group.
Because a significant portion of pacemaker patients had associated CHD, further analyses were performed comparing the portion of pacemaker patients with and without associated CHD (Figure 4A–4D). In the subgroup of pacemaker patients without CHD, PCQLI scores were generally equal to or slightly worse than patients with BAV and better than Fontan patients. In the subgroup of pacemaker patients with CHD, PCQLI scores were worse than patients with BAV and similar to Fontan patients.
Other Factors Potentially Affecting QOL
Patient age at device implant and time from initial device implant were not correlated with either patient Disease Impact or Psychosocial Impact subscale scores. There was no difference in QOL score comparing patients with or without a history of device shock for either patient or parent-proxy–reported Total PedsQL or Total PCQLI scores (data not shown). Next, we removed the patients in the ICD group with associated CHD to analyze the potential confounding effect of associated CHD within the ICD group. There was no difference in patient- or parent-proxy–reported Total PedsQL or Total PCQLI scores in patients with or without a history of device shock and no associated CHD.
Structural Equation Modeling
Drivers of Patient-Reported QOL
The presence of associated CHD and ICD device type compared with pacemaker type strongly correlated with lower QOL scores for both patient Disease as well as Psychosocial Impact subscale scores (Table 3). Issues surrounding patient self-perception, including global self-worth and athletic competence, had a strong positive correlation with lower QOL scoring. History of device complications and activity restrictions were not found to be significant drivers of QOL assessment. As a model, the total R2 for patient PCQLI Disease Impact and Psychosocial Impact subscale score models were 0.33 and 0.44, respectively.
Drivers of Parent-Proxy–Reported QOL
Lower parent-proxy Disease Impact subscale scores were seen in the presence of CHD and ICD device type (Table 4). Unlike patient reporting, in which issues surrounding self-perception seemed to be the primary driver of QOL scores, for parent-proxy scores the strongest correlation with lower QOL measurements correlated with issues surrounding patient behavior. As a model, the total R2 for parent-proxy Disease Impact and Psychosocial Impact subscale score models were 0.34 and 0.35, respectively.
Despite ongoing improvements in technology and implantation techniques, the acute and long-term management of pediatric and adolescent patients with cardiac rhythm devices continues to be associated with significant morbidity and concerns with regard to effects on QOL. In this study, we evaluated the effects of implanted cardiac rhythm devices on the QOL of pediatric and adolescent patients. Using pediatric generic and cardiac-specific measurement tools, this study demonstrated that devices significantly impacted patient- and parent-proxy–reported QOL scores. Key drivers of QOL scores included patient- and parent-specific self-perception and behavioral variables, CHD, and the presence of an ICD system as opposed to a pacing system.
Several studies have highlighted the issues surrounding the management of devices in pediatric populations.2–5 Differences in underlying cardiac disease processes, patient size, and peer-related social structures differentiate pediatric device patients from their adult counterparts. In this group, the presence of concomitant CHD was relatively high, especially in the pacemaker group, in which nearly 60% of patients had associated CHD. Similarly, the high rate of device-related complications in both device groups added additional concerns to an already high-risk group of patients. Despite these morbidities, overall measurements of self-perception, as well as emotional and behavior functioning in both device groups, were similar to previously reported measures in healthy pediatric populations.21 Despite this, the effects of devices on individual patients within these groups led to significant differences in PedsQL QOL scores compared with healthy populations and PCQLI scores in patients with mild forms of CHD, such as isolated BAV.
Similar to data from pediatric ICD studies by Demaso and Sears, the presence of an ICD was correlated with lower QOL scores in both patients and their parent-proxy. In total, the ICD group had lower PedsQL QOL scores compared with the healthy group and lower PCQLI scores compared with patients with isolated BAV. Furthermore, in terms of Disease Impact subscale scores, ICD patients have scores significantly lower than patients with TOF and similar to patients with Fontan physiology. Although the precise mechanisms behind lower QOL scores in ICD systems cannot be elucidated in this retrospective analysis, several potential causes include larger device generator, underlying disease mechanisms, and the potential for device shock.
In addition to device type, this study also underscores the synergistic detrimental effects of CHD on the QOL of patients with devices. Not surprisingly, the long-term comorbidities associated with CHD had significant additive effects associated with lower QOL, placing this vulnerable population at increased risk. This effect was most evident in the pacemaker group. Although pacemaker patients without CHD had lower PedsQL QOL scores compared with healthy controls, their PCQLI scores were comparable with patients with mild forms of CHD, such as isolated BAV. In the presence of CHD, pacemaker patients had significantly lower scores compared with the BAV group, comparable with patients with TOF and Fontan physiology.
In this pediatric cohort, there was a high rate of device shocks, and nearly 50% were inappropriate. Unfortunately, secondary to inadequate sample size, we were unable to make comparisons of QOL between patients with a history of appropriate versus inappropriate device shock. In the total ICD group, we were surprised to find that a history of any prior ICD shock was not associated with lower QOL scores. Similar to findings by Sears, history of a previous device shock was not associated with statistically lower QOL scores, although this might be affected by patient number and inadequate power for this secondary subanalysis.
When comparing patient and parent-proxy drivers of lower QOL, several similarities and differences were apparent. For both groups, the presence of CHD and ICD device type was found to be correlated with lower QOL scores. However, there were significant differences in the effects of self-perception and behavioral issues between patient and parent-proxy assessments of QOL. Although patient assessments were driven primarily by issues surrounding self-perception, parent-proxy QOL assessments were driven by behavioral issues. These differences are not unique to patients with devices, and similar findings have been demonstrated in other pediatric groups with chronic illness.22–24 These findings suggest that although patients are aware of their own self-perception, they are unaware of the outward manifestations of their behavior. In direct contrast, parents are more aware of the effects on their child’s behavior but less able to perceive the effects of their children’s own self-perception.
Secondary to the retrospective nature of the study, there was no ability to infer direct causality. This population of device patients represents only a portion of the total device patients, followed by the participating institutions. As such, potential selection bias exists and the external validity of these findings should be interpreted with this in mind. Also, although this is the largest pediatric device group reported to date, the relatively limited patient number restricted subanalysis within patient groups. Finally, for PCQLI QOL score analyses, no comparison with truly disease-free patients could be performed. Although patients with isolated BAV represent a mild form of disease within the CHD population, the knowledge of a developmentally abnormal valve and the need for chronic follow-up may potentially affect QOL and lead to an underestimate of the effect of devices if an otherwise healthy group was used.
Patient- and parent-proxy–reported QOL is significantly affected by the presence of cardiac rhythm devices and is worsened in those patients with CHD and ICD systems as opposed to pacing systems. These findings should encourage us to consider the negative impact of devices, particularly ICDs, on pediatric patients and to develop strategies to mitigate these effects. Whether these effects on QOL can be reduced through the use of psychotherapy needs to be assessed.
Ronn E. Tanel has been paid honorarium as a speaker for St. Jude fellow conference. He also has received an Institutional fellow training grant supported by Medtronic and St. Jude. The other authors have no conflicts to report.
The online-only Data Supplement is available at http://circep.ahajournals.org/lookup/suppl/doi:10.1161/CIRCEP.112.973032/-/DC1.
- Received April 5, 2012.
- Accepted October 17, 2012.
- © 2012 American Heart Association, Inc.
- Berul CI,
- Van Hare GF,
- Kertesz NJ,
- Dubin AM,
- Cecchin F,
- Collins KK,
- Cannon BC,
- Alexander ME,
- Triedman JK,
- Walsh EP,
- Friedman RA
- Duru F,
- Büchi S,
- Klaghofer R,
- Mattmann H,
- Sensky T,
- Buddeberg C,
- Candinas R
- Marino BS,
- Tomlinson RS,
- Wernovsky G,
- Drotar D,
- Newburger JW,
- Mahony L,
- Mussatto K,
- Tong E,
- Cohen M,
- Andersen C,
- Shera D,
- Khoury PR,
- Wray J,
- Gaynor JW,
- Helfaer MA,
- Kazak AE,
- Shea JA
- Uzark K,
- Jones K,
- Slusher J,
- Limbers CA,
- Burwinkle TM,
- Varni JW
- Bollen K
- Joreskog K,
- Sörbom D
- Steiger J
- Harter S
- Buyan N,
- Türkmen MA,
- Bilge I,
- Baskin E,
- Haberal M,
- Bilginer Y,
- Mir S,
- Emre S,
- Akman S,
- Ozkaya O,
- Fidan K,
- Alpay H,
- Kavukcu S,
- Sever L,
- Ozçakar ZB,
- Dogrucan N
- White-Koning M,
- Arnaud C,
- Dickinson HO,
- Thyen U,
- Beckung E,
- Fauconnier J,
- McManus V,
- Michelsen SI,
- Parkes J,
- Parkinson K,
- Schirripa G,
- Colver A
To date, there has been a paucity of data with regard to quality of life (QOL) in pediatric patients with implanted defibrillators and pacemakers. In this study, QOL was assessed using both general pediatric and cardiac-specific QOL measurement tools. It was demonstrated that defibrillators, as well as pacemakers, have significant effect on patient QOL, which is further modified by device type as well as other patient characteristics, such as congenital heart disease. Furthermore, specific psychological factors that correlate with improved or worsened QOL, as well as key differences in drivers of QOL between patients and their parents, were identified. These data should both raise awareness of this effect on QOL and help design patient-specific psychological management strategies to mitigate these deleterious effects.