INTRODUCTION
Cardiac pacemakers have become the standard of care for treating persistent, symptomatic bradyarrhythmias such as atrioventricular block, sick sinus syndrome and atrial fibrillation with bradycardia.(1) Epidemiological data on permanent pacemaker implantation (PPI) emanates predominantly from economically developed countries, with a paucity of published data from sub-Saharan Africa.(2–4) One reason for the lack of data is the high cost of pacemakers which has resulted in limited access for deserving patients in economically disadvantaged countries. Over a 10-year period between 2003 and 2013, 1257 first implant permanent pacemakers were implanted at a tertiary academic centre in Cape Town.(5) A survey in 1998 by the Cardiac Arrhythmia Society of South Africa reported an overall pacemaker implantation rate of only 39 per million in South Africa.(6) A subsequent report in 2016 showed that this rate had increased significantly to 138 per million.(7)
This study aims to describe patient demographics, clinical indications, short-term and long-term complications in patients undergoing PPI at the Charlotte Maxeke Johannesburg Academic Hospital (CMJAH), a large urban public teaching hospital in Johannesburg, South Africa. At present, approximately 100 permanent pacemakers are implanted annually in our institution. Yet complications associated with PPI have not been studied locally, despite numerous studies reporting an inverse relationship between institutional annual procedural volumes and complications.(2,8)
METHODS
We conducted a retrospective review of in-patient and out-patient medical records of patients 18 years of age and older, who underwent PPI in the division of cardiology at CMJAH between January 2009 and November 2018. We excluded patients with temporary pacemaker implantation, implantable loop recorder implants as well as patients referred for an impulse generator box change.
In our institution, active leads are placed in all patients. After pacemaker implantation or generator box change, patients are referred for a chest x-ray to exclude complications such as a pneumothorax. Cardiac resynchronization therapy (CRT) devices were not routinely implanted between January 2009 and November 2018 and this small group of patients has been excluded from the study analysis.
Data was collected from the PPI operative reports, in-patient medical records and out-patient pacemaker clinic files. Demographic data, comorbidities and chronic oral medication data were extracted. The laboratory biochemistry parameters collected included the white cell count (WCC), platelet count, international normalized ratio (INR), C-reactive protein (CRP), urea and electrolytes, creatinine and the estimated glomerular filtration rate (eGFR). The laboratory biochemistry results were retrieved from the National Health Laboratory Service electronic database.
Data from the PPI procedure included the date of the first implantation, pacemaker model, mode of pacing and the underlying cardiac rhythm. Complications for PPI were classified as short-term (occurring during the implantation admission) and long-term (occurring at any time after discharge from hospital). Complications included pacemaker implantation-related infection or sepsis, haemorrhage, haemothorax, haematoma, pneumothorax, diaphragm stimulation, lead dislodgement, lead fracture, ventricular perforation and death. For comparison purposes, we divided the study patients into two groups based on the occurrence of complications. Ethical approval for the study was obtained from the University of the Witwatersrand Human Research Ethics Committee.
Statistical analysis
Categorical variables are expressed as numbers and percentages and were compared using the Chi-square test. We used the Shapiro–Wilk test and the skewness and kurtosis test to assess for normality. Continuous variables with a normal distribution are expressed as mean and standard deviation. The median and interquartile ranges (IQRs) were used for continuous variables with a non-normal distribution. We compared normally distributed continuous variables by using the Student t-test, and the Wilcoxon rank-sum (Mann–Whitney) test was used to compare medians for non-normal data. Variables with a P-value less than 0.25 in the univariate logistic regression analysis were included in the multivariate logistic regression model. The odds ratio was calculated with its 95% confidence intervals (CI), and differences were considered statistically significant at a P-value less than 0.05. All analyses were conducted using STATA Version 16.0 (StataCorp, College Station, Texas, USA).
RESULTS
Between January 2009 and November 2018, 573 patients underwent index PPI (Figure 1). Five patients had incomplete medical records, and an additional three patients were younger than 18 years of age at the time of the first pacemaker implantation. These patients were excluded from the study (Figure 2). The final study population comprised of 565 patients, with 325 (57.5%) females. At first pacemaker implantation, the median age was 71.8 (IQR: 61.7–78.8) years, and hypertension was the most prevalent comorbidity, reported in 237 (41.95%) patients. The rest of the baseline demographic and clinical parameters are reported in Table 1.
Baseline demographic and clinical characteristics of patients referred for permanent pacemaker implantation according to the occurrence of short and long-term complications
| Variable | Overall population (n = 565) | Complications yes (n = 40) | Complications no (n = 525) | P-value |
|---|---|---|---|---|
| Age (years) | 71.8 (61.7–78.8) | 70.9 (59.9–78.1) | 71.9 (61.7–78.8) | 0.491 |
| Females | 325 (57.52) | 31 (77.50) | 294 (56.00) | 0.008 |
| Ethnicity | 0.290 | |||
| African | 260 (46.02) | 24 (60.00) | 236 (44.95) | |
| Caucasian | 227 (40.18) | 11 (27.50) | 216 (41.14) | |
| Indian | 53 (9.38) | 3 (7.50) | 50 (9.52) | |
| Mixed ancestry | 25 (4.42) | 2 (5.00) | 23 (4.38) | |
| Comorbidities | ||||
| Hypertension | 237 (41.95) | 20 (50.00) | 217 (41.33) | 0.284 |
| Diabetes Mellitus | 56 (9.91) | 5 (12.50) | 51 (9.71) | 0.570 |
| IHD | 36 (6.37) | 0 (0) | 36 (6.86) | 0.087 |
| Cardiomyopathy | 36 (6.37) | 5 (12.50) | 31 (5.90) | 0.100 |
| CKD | 24 (4.25) | 1 (2.50) | 23 (4.38) | 0.570 |
| Biochemistry | ||||
| WCC (109/L) | 7.04 (5.73–8.67) | 7.96 (5.98–9.55) | 7.01 (5.72–8.44) | 0.104 |
| Platelets (109/L) | 287 (211–366) | 262 (222–294) | 290 (209–367) | 0.258 |
| CRP (mg/L) | 10 (10–13) | 10 (10–15) | 10 (10–13) | 0.351 |
| INR | 1.07 (1.01–1.14) | 1.07 (1.04–1.21) | 1.07 (1.01–1.13) | 0.243 |
| GFR (ml/min) 1.7 m2 | 76.2 (55.3–92.4) | 76.4 (55.3–92.4) | 76.4 (55.3–92.4) | 0.489 |
Data shown as mean and standard deviation for normally distributed variables. The median and interquartile ranges (IQR) were used for continuous variables with a skewed distribution. CKD: chronic kidney disease; CRP: C-reactive protein; eGFR: estimated glomerular filtration rate; WCC: white cell count; IHD: ischaemic heart disease; INR: international normalized ratio.
Atrioventricular heart block was the most common indication for PPI and was diagnosed in 417 (73.81%) patients (Figure 3). Dual sensing, dual pacing and dual inhibition (DDD) was the most common mode of pacing (72.60%), followed by ventricular sensing, ventricular pacing and inhibition (VVI) (24.20%), atrial sensing, atrial pacing and atrial inhibition (AAI) (1.42%) and ventricular pacing, dual-sensing and dual inhibition (VDD) (1.42%). Only two patients had ventricular pacing, no sensing and no inhibition (VOO) pacemaker modes.
A total of 40 (7.08%) patients experienced short-term and long-term complications. Of the short-term complications, pacing lead dislodgement occurred in 9 (1.59%) patients, pneumothorax in 6 (1.06%) patients, infections in 3 (0.53%) patients, excessive haemorrhage in 2 (0.35%) patients and a haemothorax occurred in 1 (0.17%) patient. The exact nature of the complication was not specified in six patients. None of the patients experienced cardiac arrest peri-operatively. A 78-year-old female with a background medical history of a haematological malignancy complicated with a pneumothorax, which required an intercostal drain but further complicated with an empyema on day three post PPI. The patient was optimally treated in the intensive care unit after surgical drainage of the empyema. She later complicated with sepsis-related multiple organ failure and subsequently demised.
The most common long-term complication was lead dislodgement, which was documented in 7 (1.24%) patients, followed by sepsis in 5 (0.88%) patients, haematoma in 3 (0.52%) patients and two patients each experienced diaphragm stimulation and a lead fracture. Multivariate logistic regression analysis identified female gender as the only independent predictor for a complication, with females 3.8 times more likely to experience a complication post pacemaker implantation [odds ratio (OR): 3.80; 95% CI: 1.40–10.32, P = 0.009] (Table 2).
Univariate and multivariate logistic regression analysis for predictors of short and long-term complications
| Univariate logistic regression | Multivariate logistic regression | |||||
|---|---|---|---|---|---|---|
| Odds ratio | P-value | 95% CI | Odds ratio | P-value | 95% CI | |
| Age, years | 0.99 | 0.296 | 0.96–1.01 | |||
| Females | 2.71 | 0.010 | 1.26–5.79 | 3.80 | 0.009 | 1.40–10.32 |
| Diabetes mellitus | 1.32 | 0.571 | 0.49–3.53 | |||
| White cell count | 1.11 | 0.140 | 0.96–1.28 | |||
| Platelets | 0.99 | 0.156 | 0.99–1.00 | |||
| C-reactive protein | 1.00 | 0.770 | 0.98–1.02 | |||
DISCUSSION
In this study, symptomatic bradycardia requiring pacing was found mostly in the elderly, with AV block being the most common indication for PPI. We found that the median age at first implantation was 71.8 years, similar to data by Antonelli et al. from Israel, who reported a mean age of 74.6 years at implantation.(9)
The predominance of females in our study is possibly related to a higher life expectancy for females in South Africa.(10) In a Turkish study, nearly 50% of patients that underwent PPI were women, in keeping with our study findings.(11) In the current study, females were 3.8 times more likely to experience a complication. Similarly, Nowak et al. found a higher rate of acute complications in females referred for pacemaker implantation.(12) Moreover, in a large population-based cohort study of 28,860 Danish patients who underwent PPI, the adjusted odds ratio for pneumothorax in females was 1.9 (95% CI: 1.4–2.6, P < 0.001).(13) We are unable to hypothesize a physiologically plausible explanation for this observation. Contrary to our findings, Udo et al. reported that male gender was as an independent predictor for a complication within two months of PPI.(4)
In a multicentre database of 64,951 Japanese patients who underwent pacemaker implantation, the incidence of in-hospital complications was 2.5%.(14) The independent predictors of complications were haemodialysis, female gender, and a body mass index less than 18.5 kg/m.2(14) Kirkfeldt et al. found that implantation in a non-university hospital, an inexperienced operator who has performed less than 25 pacemaker implantations, a dual-chamber pacemaker device and passive fixation of the right atrial lead, all increased the risk of complications.(2)
Another study involving 1517 patients receiving index pacemakers between 2003 and 2007 reported a short-term and long-term complication rate of 12.4% and 9.2%, respectively. The predictors of short-term complications were male gender, advanced age, an increased body mass index, anticoagulation drugs and passive atrial lead fixation. Predictors of long-term complications were an increased body mass index, hypertension and dual-chamber device implantation.(4) These complication rates are higher than the complication rate of 7.08% found in our study cohort, but these differences may be due to diverse definitions for a clinically significant complication.
In the current study, haematology and biochemical parameters were routinely assessed in all patients before PPI. The WCC and CRP levels were done to exclude any pre-existing infective process that may predispose the patient to pacemaker sepsis.(15) Most of our study patients had WCC and CRP levels within the normal range. However, even patients with marginally elevated WCC and CRP levels did not demonstrate higher rates of sepsis compared to patients with a normal CRP and WCC.
Pacemaker sepsis is a clinically challenging complication that requires complete removal of the pacemaker device and implantation of a new device at a different site after a full course of antibiotic therapy.(16) Contemporary data indicates that the incidence of device infection after PPI is 0.1%–0.7% and 0.7%–1.2% for implantable cardiac defibrillators.(17) In our study, infection/sepsis was diagnosed in 0.53% and 0.88% of patients, in the peri-operative period and post-hospital discharge, respectively. Despite the fact that we did not document procedural times during pacemaker implantation, we hypothesize that sepsis in our patients could possibly be related to relative operator inexperience and attendant increased procedural duration as our unit is a cardiology fellow training site.
STUDY LIMITATIONS
The occurrence of complications may not have been meticulously documented in clinical notes, leading to under- reporting of complications. Another limitation of this study is the lack of data on the number of operators and the operator's level of experience as there is a known association between the occurrence of complications and the operator's level of experience. The presence of structural heart disease was also not documented, although most of our patients were routinely screened with an echocardiogram before PPI.
CONCLUSION
In our study, atrioventricular block was the most common indication for PPI. The overall institutional complication rate was 7.08%, which is very similar to rates reported in economically developed countries. Furthermore, the risk of developing a complication was found to be significantly higher in females.