© 2005 The European Society of Cardiology. Published by Elsevier Ltd. All rights reserved.
Proportionality of rate response to metabolic workload provided by a rate adaptive pacemaker with automatic rate profile optimization
aInstitute of Medicine, Department of Heart Disease, University of Bergen Bergen, Norway; bMedtronic Bakken Research Center Maastricht, The Netherlands
Manuscript submitted 17 October 2003. Accepted after revision 22 September 2004.
*Corresponding author. Institute of Medicine, Department of Heart Disease, Haukeland Hospital, 5021 Bergen, Norway. Tel.: +47 55972220; fax: +47 55975150. E-mail address: peter.schuster{at}med.uib.no (P. Schuster).
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResultsDiscussionConclusionsAcknowledgementsReferences |
|---|
OBJECTIVE: The rate response of a pacemaker (PM) was compared with the sinus rate in patients during repeated exercise tests, at different settings of the rate response parameters.
METHODS AND RESULTS: In patients with paroxysmal sick sinus syndrome (n = 3) or atrioventricular block (n = 8), a rate responsive PM was implanted. The activity-dependent pacing rate is represented by the sensor indicated rate (SIR). Each patient performed a treadmill test at 1 month, 1 year, and 2 years after implantation. Prior to the 1 and 2 year tests PM parameters were changed to produce a larger rate increase, especially at moderate levels of daily life activity. During the tests the O2 consumption and CO2 production were measured, breath-by-breath, to determine the workload and the anaerobic threshold. On average the workload (oxygen consumption), the patient's sinus rate, and the SIR, showed a linear increase with the workload imposed by the treadmill. In the 1 month and 1 year test the SIR was much lower than the spontaneous rhythm, especially at low or moderate workloads. On the more dynamic setting of several rate adaptive parameters at 2 years, the SIR changed significantly and was close to the spontaneous HR.
CONCLUSIONS: The examined PM provides a paced heart rate that is proportional to the workload. For the first time the effect of reprogramming rate response parameters to produce an SIR that is similar to the sinus rate is shown in this study.
Key Words: cardiac pacemaker, rate adaptation algorithms, activity sensor, treadmill test
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionConclusionsAcknowledgementsReferences |
|---|
Rate responsive pacemakers should provide a pacing rate that gives similar increases to that of the normal sinus node. It has been shown that adequate rate responsive pacing may enhance the exercise capability in patients with heart failure [1]
, and it appears to be safe also in patients with angina pectoris [2]. The rate response is commonly evaluated with exercise tests, and in healthy persons a linear relationship is found between HR and workload [3].
Comparison of sinus rate and the sensor indicated rate (SIR) has been made [4], but there are no studies evaluating the effect of reprogramming the PM by comparing SIR with sinus rate. Manual adjustment of the programmed parameters is considered to be useful to adapt rate modulation to the patient's needs [5,6].
Due to their age and cardiovascular condition, pacemaker patients are generally not involved in strenuous exercise. Therefore, it seems most important to ensure adequate pacing rates during normal daily activities or the SIR during the moderate activities of daily living (ADL rate). This study aimed at optimization of the ADL rate with the patients' own sinus rhythm as the reference. At different workloads we measured the workload (oxygen consumption), actual heart rate from the ECG, and the SIR from the pacemaker, during a treadmill exercise test. In subsequent tests the rate response parameters were reprogrammed for optimization.
|
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionConclusionsAcknowledgementsReferences |
|---|
Patients
The prospectively enroled study population consisted of four females and seven males (mean age 66 ± 16 years), with paroxysmal sick sinus syndrome in three and atrioventricular (AV) block in eight cases. All patients had spontaneous sinus driven heart rhythm during the tests most of the time (mean 92%), measured manually from the surface ECG. The heart rate response to exercise was analyzed separately for the patients with sick sinus and those with AV block. The rate response was very similar for the two groups and, therefore, only the combined results are presented in this paper. The study has been approved by the regional ethics committee and all the patients gave their written informed consent.
Pacemaker system
The Kappa 700 (Medtronic, Inc., Minneapolis, USA) was implanted using conventional implantation techniques with a standard unipolar lead in the right ventricle and a bipolar lead in the right atrium. The PM is a multiprogrammable dual chamber PM with rate responsive pacing capability based on a single-axis accelerometer. The PM features include rate response with automatic optimization. The PM automatically optimizes rate response by comparing the patient's current sensor rate profile in the ranges of submaximum workload (activity of daily living response) and maximum workload (exertional response) against the target rate profile, as determined by the physician. The PM target rate is dependent on the sensor assessed signal frequency and weighted by the amplitude of the sensor signal. The PM further assesses the rate response levels daily, by adjusting the amount of time the SIR is within the ADL and exertion range, based on comparing the pacemaker's historical SIR profiles against a clinician described rate profile of the patient. If the rate profiles differ, rate response is adjusted slightly (both up- and downward) in the appropriate rate range and the assessment is repeated again the next day. The independent control of rate response in both the submaximal and maximal rate ranges implies that the rate response is not necessarily linear with the workload.
The sensor used in the Kappa 700 PM responds to upright horizontal movements and determines the SIR on the basis of both signal frequency and amplitude of the previous 6 s of data. The algorithm gradually accelerates or decelerates the actual pacing rate to match the target rate during exercise.
For the programming of the PM the 9790C programmer (Medtronic Inc., Minneapolis, USA) was used.
Pacemaker programming (Table 1)
|
The automatic rate response optimization of this PM for exercise depends on the values of six programmable parameters:
- The lower rate is the lowest rate during rest, and was programmed to 60–70 min–1, which was within 10 min–1 below the instantaneous resting rate.
- The upper sensor rate is the maximum rate during vigorous activity. It was programmed to 85% of the age predicted maximum heart rate (220-age in years) as recommended [7]. For the present group of 11 patients the mean upper sensor rate was 130 min–1.
- Activities of daily living rate (ADL rate) is the rate that should be reached during moderate daily activities. The ADL rate was programmed to 95 min–1 in test 1 and 105 min–1 in tests 2 and 3.
- ADL response is the fraction of time that the SIR is at or near the ADL rate. It has a value from 1 (ADL rate 10 min/day) to 5 (ADL rate 120 min/day). The nominal value 3 was used in the first test and 4 in tests 2 and 3.
- Exertion response is the fraction of time the SIR is near the upper sensor rate. The exertion response has a value from 1(20 min/week) to 5(ca. 2 h/week). We used the value 3 in test 1, and 4 in tests 2 and 3.
- The activity threshold (low, med/low, med/high, high) indicates the exercise level at which the SIR starts to increase above the lower rate. The nominal setting med/low was programmed in tests 1 and 2 and low in test 3 to obtain a greater SIR increase at low workloads.
The first test was performed at least 28 days after implantation to allow the optimization algorithm to reach stable values for the rate responsive parameters, using the nominal PM settings from the manufacturer. For the second and third test reprogramming of the parameters occurred 14 days prior to the test, and a more dynamic response was used in terms of increasing the ADL rate and response, as well as the exertion response and decreasing the activity threshold to mimic the sinus node behaviour (Table 1). The rationale for reprogramming was based on a preliminary analysis of the SIR and to allow the automatic rate profile optimization to compare the SIR profiles against the clinician described rate profile over at least two weeks. (see Pacemaker system)
Exercise test
To assess the chronotropic response of the PM and the exercise capacity, the patients performed a symptom limited treadmill exercise test using the Minnesota Pacemaker Response Exercise Protocol (M-PREP) on the Quinton (Quinton Instrument, Co., Seattle, WA, USA) treadmill. The protocol has been designed to evaluate the performance of rate adaptive pacemakers. The protocol uses a linear increase in the workload, including typical workload ranges of activities of daily living (less than 5–6 METs) and the possibility of achieving strenuous levels of exercise by increasing speed by 0.5 mph at each stage and grade adjustments to achieve a linear increase of 1 MET per minute [8]. The metabolic unit of 1 MET is defined as the increase in workload that causes an increase of oxygen consumption of 3.5 ml kg–1 min–1, which is equal to the oxygen consumption at rest in normal subjects.
Metabolic data were collected using the breath-by-breath method, of measuring oxygen uptake (VO2), carbon dioxide output (VCO2) and minute ventilation (VE), calculating the METS from the VO2. The average METS measured over the last 16 s at every stage were used for statistical comparisons.
Data acquisition
After 5 min rest the metabolic data were collected by the Cardiorespiratory Diagnostic System CardiO2 and CPX/D utilizing BREEZEEx software (Medical graphics corporation, St. Paul, MN, USA).
The heart rate during the treadmill test was measured manually every 30 s from the recorded ECG and the automatically adjusted SIR from the PM. The SIR was transferred from the PM by telemetry at the end of the test (9790C Programmer, Medtronic Inc., Minneapolis, USA).
Data analysis
The results are presented as the mean ± standard deviation. Statistical significance between means was tested by paired t-test and a P < 0.05 was considered significant; ns = not statistically significant.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionConclusionsAcknowledgementsReferences |
|---|
Workload (Fig. 1)
|
The patients achieved the same exercise level in the three tests, and on average the walking distance was 485 ± 312 m in test 1; 441 ± 312 m in test 2; and 460 ± 310 m in test 3 (ns). The calculated maximum metabolic workload in METS compared with the actually achieved maximum workload was not significantly different in the three studies. The workload calculated from the oxygen uptake was proportional to the workload imposed by the treadmill aimed to include ADL workloads (Fig. 1). The slope of the regression line of METS versus exercise time in test 1, the intercept and the correlation coefficient were 0.71, 1.63 and 0.97, respectively. The slope of the regression line of METS versus exercise time in test 2, the intercept and the correlation coefficient were 0.77, 1.34 and 0.96, respectively. The slope of the regression line of METS versus exercise time in test 3, the intercept and the correlation coefficient were 0.71, 1.33 and 0.98, respectively.
The average maximum workload achieved during the tests was 7.2 ± 2.6 METs in test 1, 6.3 ± 2.3 METs in test 2, and 7.2 ± 2.6 METs in test 3 (ns).
Manually measured heart rate
During the tests the heart rate increased from 99 ± 15 to 139 ± 17 min–1 (test 1), 100 ± 8 to 140 ± 16 min–1 (test 2) and from 101 ± 13 to 141 ± 16 min–1 (test 3) (ns). The slope of the regression line of HR on imposed workload was 5.0 ± 97 min–1 MET–1 with a correlation of R2 = 0.92 (test 1), 5.3 ± 103 min–1 MET–1 with a correlation of R2 = 0.95 (test 2), and 5.5 ± 99 min–1 MET–1 with a correlation of R2 = 0.94 (test 3). The manually measured HR represented mostly a sinus node tracked ventricular pacing mode based on the ECG recordings, only two patients had (15–35%) atrial paced HR in a maximum of two studies. Two patients had 60 and 100% atrial paced HR in one test and were excluded from analysis in the respective tests.
|
The SIR increased during the exercise test from 59 ± 6 min–1 to 122 ± 23. The relationship between the mean SIR and the imposed workload (W) is described by: SIR = 8.9W + 50; R2 = 0.96 in test 1. In test 2 the SIR increased from 66 ± 7 min–1 to 129 ± 18: SIR = 6.4W + 79; R2 = 0.94. In test 3 the SIR increased from 68 ± 10 min–1 to a maximum of 140 ± 15 min–1, and the regression line is SIR = 5.6W + 88; R2 = 0.45. The decline of the pacing rate at stage 9 and 10 in test 2 (Fig. 2b) might be explained by the low number of patients contributing to these measurements.
Thus, the changes of the rate response parameters (Table 1) caused an increase in the SIR at all exercise levels, and the greatest increase occurred at lower exercise levels. Furthermore, in test 3 the increase in SIR with workload appears nonlinear: a steep increase occurred at workloads up to 2 METs, almost a plateau between 3 and 5 METs, and this is followed by a second steep increase to the upper sensor rate at 7–8 METs.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionConclusionsAcknowledgementsReferences |
|---|
The achieved maximum workload, the increase in sinus rate, and the O2 consumption, were on average, similar during the three exercise tests. This indicates that the condition of the patients did not change significantly in the course of 2 years, and even between tests 2 and 3, the sinus rate was similar, providing the unique possibility to compare effects on the SIR by reprogramming the device.
Patient response to exercise
The M-PREP has been designed to impose an increasing workload in equal steps of 1 MET per minute. Consistent with this design is the linear increase in the workload estimated from oxygen consumption. At low workloads the O2 consumption was higher than expected (Fig. 1), but the difference between imposed and objective workload was not significant.
The mean HR increased linearly with the workload in all the three tests, which is consistent with other studies [9]. At low workloads, however, the HR was higher than expected, e.g., about 100 min–1 at rest (Fig. 2). Thus, at low imposed workloads both the O2 consumption, and the sinus rate were higher than expected on the basis of measurements in normal subjects [1–3], and it seems that the patients perceived the workload as higher than it actually was. The underlying cause is not known, but may be related to the age and general condition of the patients, and/or it may be due to mental stress preceding the exercise tests, which may have especially influenced the first stages of the tests. Whatever the cause of this phenomenon, it has consequences for proper programming of the rate response of the pacemaker. At low workloads the SIR should be relatively high, but not at rest.
Sensor indicated rate
All the three settings of the rate response parameters (Table 1) produced a gradual increase in the SIR with the increasing workload, comparable with the SIR – workload relationship reported in earlier studies [10]. The nonlinear relationship in test 3 is due to the combination of a high value of the ADL response parameter and a low activity threshold, which causes a relatively large rate response at low workloads.
SIR versus manually measured HR
There are few studies comparing sinus rhythm with SIR [11], or PM patients with healthy subjects [12]. The present study shows that it is possible to adjust the rate response such as to resemble the natural rate response of a particular group of patients [13,14]. The SIR in test 1 was much lower than the manually measured HR, especially at the submaximal workloads (ADL), but in test 3 the two were rather close. The sinus node of the patients in the present study showed a rate response that deviated from that in healthy subjects in that the heart rate was relatively high at low workloads. This seems appropriate, however, because it coincided with relatively high oxygen consumption. Therefore, it is appropriate to adjust the rate response of the pacemaker likewise. The automatically adjusted SIR at ADL and exertion rates derived from the PM would have provided an appropriate heart rate for the patients in the third study if pacing had been required. Reprogramming the activity threshold to low in test 3 (to obtain a greater SIR increase at low workloads) showed that the exercise level based on the information of the accelerometer sensor is a major component of adequate rate response at ADL.
Limitations
The sample size might be of concern, but the time consuming test setup and the extremely long observation are the major reasons for the limited number of participants. The SIR was collected from the PM every minute, and for a more precise workload/HR relationship more frequent measurement might have been superior, but was not possible. The reprogramming of the different rate response parameters was performed after a preliminary analysis of the SIR, but was still not compared with the manually measured HR. The adjustment of the different rate response parameters was performed more dynamically during test 2 and 3, respectively to mimic the sinus node, and might have been adjusted randomly during the three studies.
|
Conclusions |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionConclusionsAcknowledgementsReferences |
|---|
The rate response of the pacemaker on average was linear with the imposed workload. To the best of our knowledge this study shows for the first time that reprogramming the rate response parameters of the pacemaker to more dynamic values resulted in significant changes in pacemaker rate response to obtain an SIR similar to that of the sinus node in this group of patients especially at low workloads. The adjustment to more dynamic programming of the examined device is recommended, the settings from study 3 (Table 1) might be used as a starting point.
|
Acknowledgements |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionConclusionsAcknowledgementsReferences |
|---|
This study was supported by the Norwegian Council on Cardiovascular Diseases.
|
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionConclusionsAcknowledgementsReferences |
|---|
[1] Greco EM, Guardini S, Citelli L. Cardiac rehabilitation in patients with rate responsive pacemakers. Pacing Clin Electrophysiol 1998; 21: 568–575.[Medline]
[2] Van Campen LC, de Cock CC, Visser FC, Visser CA. The effect of rate responsive pacing in patients with angina pectoris on the extent of ischemia on 201-thallium exercise scintigraphy. Pacing Clin Electrophysiol 2002; 25: 430–434.[Medline]
[3] Wilkoff BL and Miller RE. Exercise testing for chronotropic assessment. Cardiol Clin 1992; 10: 705–717.[Medline]
[4] Roberts DH, Baxter SE, Brennan PT, Gammage MD. Comparison of sinus node response to exercise with responses from 2 different activity-based rate-adaptive pacemakers in healthy-subjects of different age-groups. Pacing Clin Electrophysiol 1995; 18: 1882–1888.[Medline]
[5] Clémenty J, Barold SS, Garrigue S, et al. Clinical significance of multiple sensor options: rate response optimization, sensor blending, and trending. Am J Cardiol 1999; 83: D166–D171.[Medline]
[6] Lass J, Kaik J, Meigas K, Hinrikus H, Blinowska A. Evaluation of the quality of rate adaptation algorithms for cardiac pacing. Europace 2001; 3: 221–228.
[7] Carmouche DG, Bubien RS, Kay GN. The effect of maximum heart rate on oxygen kinetics and exercise performance at low and high workloads. Pacing Clin Electrophysiol 1998; 21: 679–686.[Medline]
[8] Mianulli MJ. Exercise physiology: relation to physiologic pacing. In Benditt DG (Ed.). Rate adaptive pacing 1993; Boston Blackwell Scientific Publications pp. 43–67.
[9] Treese N, MacCarter D, Akbulut O, et al. Ventilation and heart-rate response during exercise in normals-relevance for rate variable pacing. Pacing Clin Electrophysiol 1993; 16: 1693–1700.[Medline]
[10] Kay GN. Quantitation of chronotropic response – comparison of methods for rate-modulating permanent pacemakers. J Am Coll Cardiol 1992; 20: 1533–1541.[Abstract]
[11] Johnston SL, Bradding P, Watkins J. A simultaneous, noninvasive comparison with sinus rhythm, of two activity sensing, rate adaptive pacemakers, in an elderly population. Pacing Clin Electrophysiol 1991; 14: 20–27.[Medline]
[12] Sulke AN, Pipilis A, Henderson RA, Bucknall CA, Sowton E. Comparison of the normal sinus node with 7 types of rate responsive pacemaker during everyday activity. Br Heart J 1990; 64: 25–31.
[13] Celiker A, Ceviz N, Alehan D, Lenk MK, Ozme S. Comparison of normal sinus rhythm and pacing rate in children with minute ventilation single chamber rate adaptive permanent pacemakers. Pacing Clin Electrophysiol 1998; 21: 2100–2104.[CrossRef][Medline]
[14] Lau CP, Leung SK, Guerola M, Crijns HJ. Comparison of continuously recorded sensor and sinus rates during daily life activities and standardized exercise testing: efficacy of automatically optimized rate adaptive dual sensor pacing to simulate sinus rhythm. Pacing Clin Electrophysiol 1996; 19: 1672–1677.[Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||