Risk of pacemaker implantation after femur fracture in patients with and without a history syncope: a Danish nationwide registry-based follow-up study

Sara LK Clemmensen ?,Kristian Kragholm2 ,Bhupendar Tayal ,Christian Torp-Pedersen3 ,S?ren Kold ,Peter S?gaard5 ,Sam Riahi5

1.Department of Cardiology,Aalborg University Hospital,Aalborg,Denmark;2.Unit of Clinical Biostatistics and Epidemiology,Aalborg University Hospital,Aalborg,Denmark;3.Department of Clinical Epidemiology,Aalborg University Hospital,Aalborg,Denmark;4.Department of Orthopedics,Aalborg University Hospital,Aalborg,Denmark;5.Department of Clinical Institute,Aalborg University Hospital,Aalborg,Denmark

ABSTRACT BACKGROUND It has previously been described that fall-associated injuries including fractures are commonly observed among patients with bradyarrhythmia.However,knowledge on the risk of pacemaker implantation after admission due to femur fracture from large population-based epidemiologic data is lacking.Therefore,we investigated the risk of pacemaker implantation following femur fracture in patients with and without a history of previous syncope.METHODS All patients with femur fracture between 2005-2017 were identified using the Danish Nationwide Patient Registry.Among these,patients already having a pacemaker were excluded.Primary outcome was one-year risk of pacemaker implantation and secondary outcome was one-year all-cause mortality.Multivariable logistic regression was used to obtain absolute and relative risks of the study endpoint in relation to patients with versus without history of syncope and standardized to the age,sex,selected comorbidity and pharmacotherapy distribution of all patients.RESULTS Of 93,093 patients with femur fracture,5508 (5.9%) had a history of syncope within five years.Patients with prior syncope were slightly older (84 vs.83 years),more often male (33.6% vs.29.4%),and had more often comorbidities relative to those without history of syncope.All-cause mortality was significantly higher among those with previous history of syncope compared to those without previous syncope (29.9% vs.28.6%,P=0.021).The relative mortality risk was 1.05 (95% CI: 1.01-1.09,P=0.021).A total of 695 (0.8%) patients underwent pacemaker implantation within 5 years following femur fracture,and a significantly higher proportion of patients with syncope had a pacemaker implanted within one year (1.6% vs.0.7%,P <0.001;relative risk,2.01 [95% CI: 1.55-2.46]).CONCLUSIONS In patients with femur fracture,a history of syncope was significantly associated with a higher one-year risk of pacemaker implantation.

Syncope and falls are interrelated and common conditions causing substantial morbidity and mortality.[1]Distinction between syncope and falls can at times be difficult and a similar underlying pathophysiology is plausible.[2,3]Falls and syncope,in particular among the elderly,are associated with injuries,[1,4]including femur fractures,which is associated with not only a raised risk of long term rehabilitation but also up to 20% risk of death.[5,6]

As a result of the aging general population,the number of fall-related contacts with the health system has been increasing.[7]Furthermore,there is evidence showing that falls associated with syncope are the major cause of femur fracture in elderly.[8,9]

Cardiac disease is an underlying risk factor for syncopal events,in particular the more severe types of syncope[1,10-13]and the association between cardiac disease and falls is shown to be strong,[3,10,11,14-17]and increases with age.[12,14]Studies have shown that syncope events are significantly reduced after insertion of a pacemaker,and it further reduces the frequency of falls and fractures and injuries associated with falls.[15,18-21]Nonetheless,there is limited knowledge on fall-and syncope-related femur fractures caused by bradycardia leading to pacemaker implantation.

The aim of this study was to examine whether a history of previous syncope,in this study 5 years prior,to femur fracture in older patients is an important risk factor for a subsequent pacemaker implantation.Therefore,using nationwide administrative data on patients hospitalized due to first-time femur fracture,we investigated the absolute and relative oneyear risks of pacemaker implantation and all-cause mortality in patients with and without a history of syncope at the time of hospitalization due to femur fracture.

METHODS

Study Population and Setting

From the Danish National Patient Register,all patients admitted with a femur fracture diagnosis in the period from 2005 to 2017 in Denmark were identified.Of these,patients who already had a pacemaker and patients <65 years old were excluded.

Data Sources and Variables

All citizens in Denmark are given a unique personal number,which makes identification in and linkage between national administrative registries possible.This unique number along with date of birth,death and gender of all residents are listed in the Danish Civil Registration System.[22]The Danish National Patient Register (DNPR) contains information on diagnosis and procedure codes of all in-patient contacts since 1977 and all out-patient contacts since 1995.[23]The positive predictive value of a code for proximal femur fracture in the DNPR is 97%.[24]The diagnosis and surgical procedures are coded based on the World Health Organization’s International Classification of Diseases version 8 (ICD-8)before 1994 and version 10 (ICD-10) since 1994.The DNPR was used to assess certain comorbidities within the last five year prior to the femur fracture hospitalization including hypertension,diabetes,chronic kidney disease (CKD),ischemic heart disease (IHD),prior myocardial infarction (MI),chronic obstructive pulmonary disease (COPD),chronic heart failure (CHF) and atrial fibrillation or flutter(AF).In addition,for diabetes,COPD and hypertension,we furthermore used relevant prescription medication data from the Danish National Prescription Registry,which contains all dispensed prescription drugs in Denmark since 1995.[25]For diabetes and COPD,we assessed antidiabetic and specific inhalation drugs within the last 180 days up to the femur fracture hospitalization.For hypertension,we assessed a combination of at least two antihypertensive drugs in two consecutive quarters prior to the femur fracture hospitalization,as done and previously validated.[26,27]The Danish National Prescription Registry was furthermore used to determine the patients with prescriptions for selected medications (beta blockers,verapamil,digoxin,and amiodarone).

Main Exposure Variable

The patients were divided into two groups based on whether the patients had a history of syncope within the last 5 years prior to the femur fracture or not to compare risk between these two groups.

End-point

The main end point of this study was a one-year risk of pacemaker implantation following femur fracture.One-year all-cause mortality was a secondary endpoint.

Statistical Analysis

Continuous variables were presented as median with 1st-3rdquartiles (Q1-Q3: 25%-75% percentiles)and compared using the Wilcoxon Mann-Whitney U test,whereas categorical variables were presented in absolute numbers as well as percentages and compared using the Pearson’s Chi-square test.Univariable and multivariable Cox regression was performed to obtain hazard ratios with 95% intervals.Further,multivariable Cox regression was performed to obtain absolute and relative risks of the oneyear study endpoints in patients with versus without a history of syncope,standardized to the age,sex,comorbidity and selected pharmacotherapy distribution of all patients.Standardization was used to ensure that patients with and without a history of syncope had similar age,sex,comorbidity,and pharmacotherapy distributions.Specifically,the following covariates were included: age (categorized into groups of 65-74,75-84 and ≥ 85 years),sex,hypertension,diabetes,CKD,IHD,prior MI,COPD,CHF,AF and medication therapy with betablocker therapy,calcium antagonist,digoxin or amiodarone.Data management and statistical analyses were performed using SAS version 9.4 and R,version 3.6.1.[28]

Ethics

In Denmark,ethical approval is not required for register-based studies.The study was approved by the Data responsible unit in the Capital Region of Denmark (P-2019-404).

RESULTS

Patients

Between 2005 and 2017,119,618 patients were admitted with femur fracture.After excluding 22,439 patients under the age of 65 years and 4,086 patients with a pacemaker implanted prior to femur fracture,93,093 patients formed the final study population.All patients were divided into two groups,one group with previous syncope (n=5,508) and one without previous syncope (n=87,585).The median age for patients with syncope was 84 years (Q1-Q3: 78-89) and 33.6% of them were males,whereas patients without syncope had a median age of 83 years (Q1-Q3: 76-88) and 29.4% of them were males.

Characteristics

The baseline characteristics of the study population is listed in Table 1.Patients with previous syncope had a higher frequency of comorbidities compared to patients without previous syncope.There was a higher prevalence of hypertension (68.9%vs.55.1%),diabetes (5.5%vs.4.4%),CKD (3.0%vs.1.9%),IHD (19.8%vs.11.6%),prior MI (7.8%vs.4.5%),COPD (5.2%vs.4.1%),CHF (9.1%vs.6.1%) and AF(9.7%vs.5.8%).This was also apparent in the use of rate-or rhythm control pharmacotherapy,wherepatients with previous syncope more often were prescribed beta blockers (6.4%vs.4.2%),digoxin (1.9%vs.1.4%) and amiodarone (0.6%vs.0.2%).For verapamil,the difference was however insignificant (0.5% in patients with prior syncopevs.0.5%in patients without a history of syncope).

Table 1 Baseline characteristics.

Pacemaker Risk

At the end of the one-year follow-up period,695(0.8%) patients had a pacemaker implanted.The standardized absolute risk of patients ending up with a pacemaker was higher among patients with previous syncope (1.4%vs.0.7%,P<0.001) (Figure 1).The unadjusted and adjusted hazard ratios as well as the standardized relative risk of getting a pacemaker implantation among patients with syncope is shown in Figure 2.

Figure 1 Cumulative incidence of pacemaker implantation in the follow-up period of 12 months among the two groups,with and without previous syncope,with a significant difference between the two throughout the follow-up period.

All-cause Mortality Risk

The standardized one-year all-cause mortality risk was higher among patients with previous syncope in comparison to those without previous syncope (29.9%vs28.6%,P=0.021;Figure 3).The standardized relative risk was 1.05 (95% CI: 1.01-1.09,P=0.021).

Figure 2 The figure illustrates the unadjusted,adjusted and standardized 1-year risk ratio for pacemaker implantation and mortality.

Figure 3 Cumulative proportion of mortality in the follow-up period of 12 months among the two groups,with and without previous syncope.

DISCUSSION

In this large nation-wide registry-based study of 93,093 patients above 65-years of age with a femur fracture investigating the subsequent one-year risk of pacemaker implantation in cases with versus without a history of syncope prior to hospitalization due to femur fracture,the average risk of getting a pacemaker within the first year after femur fracture was overall low at 0.8% but twice as high in patients with a history of syncope compared to patients without a history of syncope.Furthermore,the oneyear mortality was higher among patients with previous syncope.

The connection between syncope and falls/fallrelated injuries is well-known,and this association is even more evident in the elderly.[1,4,12,14]Patients sustaining a fragility fracture have a significant higher risk of mortality,[29]why we included patients above 65-years of age.The overall one-year mortality has been found to be high both for proximal femoral fractures classified as hip fractures[30]and for distal femoral fractures.[31]This current study investigated the connection between femur fracture,a history of syncope and the subsequent risk of pacemaker implantation.The current study is the first,to our knowledge,to investigate this specific issue on a large-scale population.

The relation between syncope and fall-related injuries including fractures is well-described.[32,33]It could be a possible explanation for the connection between syncope,fracture and pacemaker implantation found in the current study,in which a twofold higher risk of pacemaker implantation in patients with a history of syncope prior to femur fracture hospitalization was seen when compared to patients without a history of syncope prior to femur fracture.

Other studies have shown syncope to be related to a worse prognosis with a higher mortality than those without syncope,[13,34]especially syncope of cardiac origin.[13,35]In our study,the risk of all-cause mortality was similarly slightly higher in patients with a history of syncope,which also indicates a more severe prognosis in patients with femur fracture with a prior history of syncope.

In our study,the patients with prior syncope were more often men,older and had numerous comorbidities.These findings are equivalent with other studies,and cardiac syncope is more frequently seen in the elderly.[14]A recent study by Probst,et al.,[36]found comorbidities,older age and male sex to be linked to a more serious outcome following syncope.However,some studies found syncope to be more prevalent in women.[37,38]Thus,these findings along with the finding of our study indicate a motive for a closer monitoring and follow-up of these high risk-patients.

A study by Mills,et al.,[39]did not find screening with a Holter-monitor justifiable in the search for arrythmia as a cause of femur fractures.This is coherent with findings of Taylor and Stout.[40]However,Mills,et al.,[39]found Holter monitoring justified in specific cases,including sign of conduction defects,bradycardia and PR-prolongation in the electrocardiogram (ECG).

The current study did not investigate the patient’s ECG’s as they were not available for interpretation;however,addition of patient’s ECG’s could be an interesting angle for future studies of risk stratification.Another noteworthy aim for forthcoming studies could be other risk factors for pacemaker implantation,as these were out of scope for this study.

Another interesting study would be to examine whether the risk of getting a pacemaker is higher in elderly patients with syncope and femur fracture than in patients with syncope but without such fracture.This is however out of scope for this current study but warrants further investigation.

Altogether,our study shows that the risk of pacemaker implantation after femur fracture was significantly higher in patients with a prior history of syncope and that such patients constitute an important population for clinical risk stratification and awareness.

LIMITATIONS

Our study has some limitations.First,it is observational of nature,meaning that the results represent associations and may not be causal.However,the use of standardization ensured comparability between patients with versus without a history of syncope.As such,the association between a history of syncope and the approximately two-fold increased risk of pacemaker implantation among patients with femur fracture may therefore likely represent a true causal relation but cannot be certainly concluded due to the observational design of our study.Second,a number of important clinical variables were not readily assessable including ECG parameters.As noted above,future examination of selected ECG parameters and signs in relation to this study scope is warranted,including the final diagnosis for pacemaker implantation as an interesting aspect.Furthermore,the reasons for syncope in patients not treated with pacemaker implantation were not assessed in this study,for this would be too extensive.Third,causes of mortality,including plausible arrhythmia-related causes,are subject to a large degree of uncertainty and therefore,we were unable to report these.

CONCLUSIONS

In this large nationwide,register-based study of patients above 65-years of age hospitalized due to femur fracture,the risk of getting a pacemaker within the first year after admission was significantly elevated among patients with a history of syncope,compared with those without such history.These results may have potential implications on cardiac monitoring of patients with femur fracture and previous history of syncope.