Ablation therapy following unsuccessful electrical cardioversion in patients with persistent atrial fibrillation

Abstract

Electrical cardioversion (ECV) a widely utilized intervention for persistent atrial fibrillation (AF) aimed at restoring sinus rhythm. However, ECV can be ineffective, raising questions about subsequent treatment options. This study aimed to compare the outcomes of non-ablation therapy versus ablation therapy following unsuccessful ECV. A total of 125 consecutive patients with persistent AF who underwent unsuccessful ECV between November 2017 and August 2023 was included in this retrospective analysis. Of these, 51.2% received only medical therapy (non-ablation therapy group, n = 64), while 48.8% underwent AF ablation (ablation therapy group, n = 61). Various ablation methods were employed, including catheter and thoracoscopic ablation. Ablation therapy was associated with significantly better AF-free survival compared to non-ablation therapy [hazard ratio (HR), 0.37; 95% confidence interval (CI) 0.22–0.61; p < 0.01]. There was no difference of AF-free survival between catheter ablation and thoracoscopic ablation groups (HR 0.79, 95% CI 0.34–1.83; p = 0.58). AF duration > 5 year (HR 1.51; 95% CI 0.930–2.437; p = 0.10), BMI ≤ 25 kg/m² (HR 1.61; 95% CI 1.004–2.581; p = 0.05) and diabetes (HR 2.38; 95% CI 0.902–6.266; p = 0.08) were considerable as predictor of AF recurrence. Ablation therapy following unsuccessful ECV was associated with maintaining sinus rhythm, regardless of the specific ablation method utilized.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-024-73989-2.

Introduction

Atrial fibrillation (AF) is a common cardiac arrhythmia leading to potential complications such as stroke and heart failure (HF). Electrical cardioversion (ECV) is an effective intervention to restore sinus rhythm in patients with persistent AF. Nonetheless, 10.3% of patients with recent onset AF did not respond to ECV¹.

There are some studies about predictors of unsuccessful ECV in AF^2–5. In early post-ablation ECV, one study reported that early post-ablation ECV failure was not associated with AF recurrence⁶, while another study concluded that it was associated⁷. There is only one article in which patients with unsuccessful ECV reportedly experienced more frequent recurrence of AF after ablation than those with successful ECV⁸. Consequently, the treatment management following unsuccessful ECV in AF patients remains an area of interest and clinical significance.

Our study aims to address this gap in knowledge by analyzing the differential outcomes between non-ablation and ablation therapies following unsuccessful ECV in patients with persistent AF. By assessing the efficacy of these treatments, the study seeks to provide insights into the optimal management approach for this patient population. Furthermore, the study sheds light on the demographic and clinical factors that predict AF recurrence following ablation therapy, such as age and AF duration, and their influence on treatment outcomes.

Results

Baseline characteristics

Baseline characteristics were analyzed and compared between the non-ablation therapy group (n = 64) and the ablation therapy group (n = 61), as shown in Table 1. All patients received oral anticoagulation therapy for at least 3 weeks prior to the ECV^9,10. Among patients with persistent AF who underwent ECV, 8% experienced unsuccessful ECV (Fig. 1). The mean durations of AF and follow-up periods were 47.6 months and 27.0 months, respectively. The mean age at the time of treatment with ECV was 61.7 years, with 28 (22%) female participants. Significantly different characteristics were observed between the two groups, including age, male, body mass index (BMI), hypertension, diabetes prevalence, cardiac implantable electronic device (CIED), and serum hemoglobin concentration. In our analysis, there was an association between age and AF duration (Supplementary Fig. 1) and comparable AUC value of these factors on prediction of AF recurrence (Supplementary Fig. 2). The observed differences in serum hemoglobin concentration between the groups may be influenced by age-related factors¹¹.

Table 1.

Baseline characteristics.

	Total (n = 125)	Non-ablation therapy (n = 64)	Ablation therapy (n = 61)	p-value
Age (years)	61.7 ± 11.2	66.2 ± 11.5	57.0 ± 8.6	< 0.01
Male, n (%)	97 (78)	42 (66)	55 (90)	< 0.01
BMI (kg/m2)	25.6 ± 3.2	24.9 ± 3.4	26.3 ± 2.9	0.02
AF duration (months)*	14.5 (5.0–75.8)	19.0 (6.0–65.0)	12.5 (4.0–78.5)	0.63
Hypertension, n (%)	60 (48)	39 (61)	21 (34)	< 0.01
Diabetes, n (%)	23 (18)	18 (28)	5 (8)	< 0.01
CAD, n (%)	11 (9)	7 (11)	4 (7)	0.39
Stroke, n (%)	6 (5)	4 (6)	2 (3)	0.44
CKD, n (%)	2 (2)	2 (3)	0 (0)	0.16
CIED, n (%)	8 (6)	7 (11)	1 (2)	0.03
AAD, n (%)	72 (58)	32 (50)	40 (66)	0.10
Digoxin, n (%)	15 (12)	10 (16)	5 (8)	0.20
RASi, n (%)	42 (34)	21 (33)	22 (36)	0.70
Beta blocker, n (%)	46 (37)	26 (41)	20 (33)	0.36
SGLT2i, n (%)	7 (6)	3 (5)	4 (7)	0.65
Statin, n (%)	34 (27)	14 (22)	20 (33)	0.17
Hb (g/dl)*	14.0 (13.0–15.0)	13.0 (12.0–14.0)	15.0 (14.0–16.0)	< 0.01
GFR (mL/min/1.73 m2)	84.6 ± 15.4	83.4 ± 15.9	85.8 ± 14.8	0.42
LVEF (%)	56.1 ± 9.7	55.6 ± 10.3	56.6 ± 9.1	0.56
Left atrium (mm)	48.4 ± 7.1	48.7 ± 7.5	48.0 ± 6.6	0.58
LAVI (mL/m2)a	49.5 (41.0–60.0)	55.5 (46.3–70.8)	46.5 (39.3–57.8)	< 0.01
E/e′	10.1 ± 4.4	11.3 ± 5.0	9.0 ± 3.4	0.01
RVSP (mmHg)*	28.0 (23.0–32.0)	30.0 (25.0-35.8)	26.0 (22.0–30.0)	< 0.01

^aThese variables exhibited a non-normal distribution and was reported as medians with interquartile ranges. AAD anti-arrhythmic drugs, BMI body mass index, CAD coronary artery disease, CIED cardiac implantable electronic device, CKD chronic kidney disease, E/e′ ratio of mitral peak E velocity to tissue Doppler early diastolic velocity e′, Hb hemoglobin, GFR glomerular filtration rate, LAVI left atrium volume index, LVEF left ventricular ejection fraction, RASi renin angiotensin system inhibitor, SGLT2i sodium glucose cotransporter-2 inhibitor, RVSP right ventricular systolic pressure.

Fig. 1.

Study flowchart. Out of the 1567 patients with persistent atrial fibrillation and complete records, 125 (8.0%) experienced unsuccessful ECV. Among them, 64 received only medical therapy without ablation therapy, while 61 received both ablation therapy and medical therapy.

The mean left ventricular ejection fraction (LV EF) was 56.1 ± 9.7%. Patients exhibited enlarged left atrial (LA) size (mean, 48.4 ± 7.1 mm), elevated left atrial volume index [LAVI; median, 49.5 mL/m²; interquartile range (IQR) 41.6 to 60.0], increased ratio of mitral peak E velocity to tissue Doppler early diastolic velocity (E/e′; mean: 10.1 ± 4.4), and higher right ventricular systolic pressure (RVSP; median, 28.0 mmHg; IQR, 23.0 to 32.0) compared to normal ranges. These findings are consistent with the association between persistent AF and HF with preserved ejection fraction. Upon comparison between the two groups, the non-ablation therapy group exhibited a higher E/e′ ratio (11.3 ± 5.0 versus 9.0 ± 3.4, p = 0.01) and RVSP (median 30.0, IQR 25.0 to 35.8 versus median 26.0; IQR 22.0–30.0, p < 0.01) and a trend toward larger LAVI (median 55.5; IQR 46.3 to 70.8 versus median 46.5; IQR 39.3 to 57.8 mL/m², p < 0.01) compared to the ablation therapy group.

Follow up period and waiting time for ablation

The mean duration of the follow-up period was 27.0 ± 19.9 months (median, 26 months; IQR, 9.0 to 38.5) in total cohort, 28.5 ± 22.2 months (median, 27.5 months; IQR 9.0 to 45.8) in the non-ablation therapy group and 25.3 ± 17.2 months (median, 25.0 months; IQR 9.5 to 36.0) in the ablation therapy group. There was no difference of follow up duration between these groups (p = 0.37).

Of the 61 patients who were assigned to the ablation therapy group, 30 patients (49%) received the catheter ablation, while 31 patients (51%) underwent the thoracoscopic ablation. Of the 31 patients who underwent thoracoscopic ablation, 11 also underwent catheter ablation. Mean waiting time for ablation was 8.8 ± 7.8 months (median, 5.0 months; IQR 3.0 to 13.0). There was a no difference of waiting time for ablation between catheter ablation group and thoracoscopic ablation group (p = 0.33, Supplementary Table 1).

Outcome

At the end of the retrospective observation, recurrence of AF was documented in 56% (n = 70) of the total cohort (non-ablation therapy group: 73%, n = 47 versus ablation therapy group: 38%, n = 23; p < 0.01). Patients in the ablation therapy group demonstrated significantly better AF-free survival compared to those in the non-ablation therapy group. The Kaplan–Meier analysis showed that the rate of the AF recurrence was significantly lower in the ablation therapy group than in the non-ablation therapy group [hazard ratio (HR), 0.37; 95% confidence interval (CI), 0.22–0.61; p < 0.01 by Cox regression, Fig. 2.] and there was no statistically significant difference of AF recurrence between catheter ablation and surgical ablation subgroup (HR 0.79; 95% CI 0.34–1.83; p = 0.58 by Cox regression, Supplementary Fig. 3). There was no significant difference of complication incidence including pericardial effusion, pacemaker, heart failure and composite outcome (Table 2.).

Fig. 2.

Kaplan–Meier curves for the atrial fibrillation-free survival. For comparison, the orange and blue lines represent the AF-free survival values after ECV in the non-ablation therapy group (n = 64) and ablation therapy group (n = 61), respectively.

Table 2.

Complication incidence of non-ablation therapy and ablation therapy groups.

	Non-ablation therapy (n = 64)	Ablation therapy (n = 61)	Odds ratio (95% CI)	p-value
Pericardial effusion	2 (3)	1 (2)	0.52 (0.046–5.849)	0.59
Pacemaker	4 (6)	1 (2)	0.25 (0.027–2.303)	0.22
Heart failure	5 (8)	7 (11)	1.53 (0.458–5.107)	0.49
Stroke	0 (0)	0 (0)
Composite outcomea	11 (17)	9 (15)	0.83 (0.319–2.179)	0.71

^aComposite outcome means sum of incidences of pericardial effusion, pacemaker, heart failure and stroke.

Predictors of atrial fibrillation recurrence

Table 3. displays the results of both univariate and multivariate analyses. We focused on factors that exhibited significant differences between the non-ablation and ablation groups: age, male, BMI, AF duration, hypertension, diabetes, CIED, serum hemoglobin level, LAVI and the E/e’ ratio. In the analyses, ‘non-ablation’ served as the reference category for ablation, and ‘none’ for diabetes. In univariate analysis, BMI > 25 kg/m² (HR 0.62; 95% CI 0.387–0.996; p = 0.05), Diabetes (HR 2.38; 95% CI 0.902–6.266; p = 0.08), and AF duration > 5 years (HR 1.51; 95% CI 0.930–2.437; p = 0.10) emerged as considerable predictors of AF recurrence after ECV. However, in multivariate analysis, only ablation therapy remained a significant predictor of AF recurrence (HR 0.29; 95% CI 0.151–0.543; p < 0.01).

Table 3.

Univariate and multivariate analyses for predictors of atrial fibrillation recurrence.

	Univariate HR (95% CI)	p-value	Multivariate HR (95% CI)	p-value
Ablation	0.37 (0.222–0.611)	< 0.01	0.29 (0.151–0.543)	< 0.01
Age > 65	1.31 (0.812–2.115)	0.27	0.78 (0.411–1.477)	0.44
Male	0.79 (0.437–1.411)	0.42	0.60 (0.277–1.287)	0.19
BMI > 25 kg/m2	0.62 (0.387–0.996)	0.05	0.85 (0.485–1.473)	0.55
AF duration > 5 years	1.51 (0.930–2.437)	0.10	1.56 (0.881–2.753)	0.13
Hypertension	1.18 (0.738–1.886)	0.49	0.87 (0.498–1.503)	0.61
Diabetes	2.38 (0.902–6.266)	0.08	1.18 (0.583–2.406)	0.64
CIED	1.27 (0.547–2.927)	0.58	0.83 (0.272–2.549)	0.75
Hb < 12 mg/dl	0.59 (0.268–1.281)	0.18	0.55 (0.233–1.299)	0.17
LAVI > 60 mL/m2	1.31 (0.781–2.203)	0.31	1.03 (0.550–1.932)	0.92
E/e′ >9	0.87 (0.531–1.437)	0.59	0.95 (0.507–1.784)	0.88

BMI body mass index, CIED cardiac implantable electronic device E/e′, ratio of mitral peak E velocity to tissue Doppler early diastolic velocity e′, LAVI left atrium volume index.

Discussion

To our knowledge, this is the first study to report an outcome of the ablation strategy following ECV failure in persistent AF patients. Ebert M, et al.⁶ documented an outcome of ECV failure soon after ablation. Kamada H, et al.⁸ compared post-ablation outcomes between successful and unsuccessful ECV groups but did not show comparison between pharmacological only therapy and ablation therapy. Therefore, the effectiveness of ablation in patients with persistent AF after unsuccessful ECV remains largely unknown. Our study revealed that, despite unsuccessful ECV, ablation therapy is associated with better AF-free survival compared to pharmacological therapy alone. Notably, AF duration emerged as the predictor of AF recurrence, with no significant differences observed in AF-free survival among various ablation methods. These findings align with those of Ebert M, et al.⁶ and Richter B, et al.¹², who similarly demonstrated that unsuccessful ECV is not necessarily associated with long-term AF recurrence. These observations challenge the assumption that unsuccessful ECV inevitably leads to permanent AF.

Ablation therapy demonstrated a significant association with AF-free survival, irrespective of AF type^13,14. This aligns with previous evidence suggesting that catheter ablation may reduce the risk of death from any cause in HF patients, regardless of severity^15–17. However, the literature on the outcomes of catheter ablation after unsuccessful ECV is limited and somewhat contradictory. Several studies have suggested that unsuccessful ECV may have implications for long-term outcomes and could potentially render patients unsuitable candidates for catheter ablation. For instance, Nakamaru et al.⁷ reported that failed ECV in patients with early recurrence of atrial arrhythmia within a blanking period of 3 months after catheter ablation was an independent predictor of recurrence, regardless of AF type. Similarly, Kamada et al.⁸ found that the AF-free survival rate was significantly higher in the successful ECV group compared to the unsuccessful ECV group in patients undergoing catheter ablation after ECV, which contrasts with our results.

Several factors may contribute to the differences observed between these studies and ours. Firstly, our study included patients undergoing two or more methods of ablation, such as catheter and thoracoscopic ablations, whereas previous studies primarily enrolled patients undergoing radiofrequency ablation or cryoablation but not thoracoscopic ablation. In our study, thoracoscopic ablation showed comparable rhythm outcome (Supplementary Fig. 2). Consequently, our study involved a more aggressive ablation therapy approach compared to other studies. Second, previous studies typically compared outcomes between successful and unsuccessful ECV groups in patients undergoing ablation therapy rather than comparing only a pharmacological therapy group and an ablation group in patients having undergone unsuccessful ECV, as was done in our study. This difference in comparison targets could contribute to variations in observed outcomes. Last, the sample size of the ablation group of unsuccessful ECV patients in the study by Kamada H, et al. was relatively small (n = 34), which may have limited its ability to identify statistical significance in outcomes. In summary, our findings underscore the potential therapeutic benefit of aggressive ablation therapy in patients that have undergone unsuccessful ECV.

There was no statistically significant difference in clinical outcomes, including unmasked sick sinus syndrome requiring cardiac implantable electronic devices, stroke, or heart failure, between the two groups (Table 2.). Ablation therapy did not reduce the aforementioned complications related to AF compared to non-ablation therapy. We tend to focus on the rhythm outcome of AF, but it is also important to evaluate clinical outcomes that impact the quality of life and long-term survival of the patients.

BMI > 25 kg/m² was associated with lower AF recurrence in univariate analysis, not in multivariate analysis. After AF ablation, recurrence rates of AF increased incrementally with increasing BMI. However, in situation of the BMI < 20 kg/m², recurrence rates of AF increased incrementally with decreasing BMI¹⁸. The predictive value of the lower figure of BMI in our study reflected the result of the previous study. Our study reiterates the importance of AF duration in prior therapy but there is no significant in multivariate analysis. Long-standing AF without ablation therapy is associated with atrial remodeling, increased AF burden, late recurrence^19,20, and poor prognosis²¹. AF duration prior to ablation is a significant independent predictor for recurrence of AF^6,8. Although the overall prevalence of AF in the United States is 1–2%²², AF duration is easily underestimated since about one-third of the total AF population is asymptomatic²³. Early detection of AF, particularly through the use of wearable smart devices²⁴, may be instrumental in preventing late recurrence and improving overall outcomes. The protective effect of early rhythm control for cardiovascular events exhibited a linear decrease with age²⁵. The efficacy of the surgical ablation was also worse in elderly patients (age ≥ 75 years)²⁶. While diabetes is known to predict AF recurrence after ablation, its significance in persistent AF remains less pronounced²⁷. In our study, diabetes showed unfavorable trends in univariate analysis but did not reach statistical significance in multivariate analysis, suggesting limited effects in persistent AF compared to paroxysmal AF.

Our study did not find significant differences in AF-free survival among various ablation methods. A randomized controlled trials have demonstrated the non-inferiority of catheter ablation compared to surgical ablation in long-standing persistent AF²⁸. Therefore, ablation therapies including catheter and thoracoscopic ablations are effective in patients with persistent AF following unsuccessful ECV.

The main limitation of our study stems from its retrospective nature, relying on chart reviews for data collection. Not all patients with persistent AF received ECV consistently due to contraindications, such as detected thrombus, or refusal of a procedure. Although we routinely attempted electrical cardioversion before ablation procedures and performed ECV for all enrolled patients in the study, there is a potential selection bias. Second, there may be a selection bias in the ablation therapy group, as patients opting for this treatment may come from better socioeconomic status and compliance compared to those in the non-ablation group, leading to potential overestimate of the efficacy of ablation therapy due to confounding variables. Third, the method used to determine AF duration might have led to underestimation, as it was based on the date of the first documented AF electrocardiogram. The integration of smart devices in future studies could provide more accurate insights into AF episode occurrence. Fourth, the use of low-escalating energy shocks for ECV in our center might have rendered the procedure less efficient compared to protocols using maximum-fixed shocks²⁹, potentially overestimating the number of patients experiencing unsuccessful ECV. Fifth, surgical ablation, unlike catheter ablation, may have various factors that could influence postoperative recurrence. Therefore, a more accurate comparison may be achieved by comparing only the groups that underwent catheter ablation with those that did not, excluding surgical ablation. Last, the inclusion of Kaplan-Meier curves for non-ablation therapy could be inappropriate because the baseline rhythm for the non-ablation therapy group is likely to be AF. Patients who do not respond to ECV and only receive drug treatment may still restore or maintain sinus rhythm. Similarly, patients who undergo ablation therapy can also continue to have AF or experience early recurrence. Therefore, previous studies have included patients in the non-ablation group as survivors regardless of their initial rhythm. Some studies have established a blanket period for the non-ablation group as well, initially including all patients as survivors and then identifying AF occurring after the blanket period as an event. We did not apply a blanket period in our study because there were no cases in our study population where AF during the blanket period led to conflicting results, regardless of whether the patients were in the ablation or non-ablation group.

Conclusion

The study contributes insights into the management of persistent AF following unsuccessful ECV, highlighting the efficacy of AF ablation including thoracoscopic ablation as a therapeutic option. These findings hold implications for optimizing clinical decision-making and enhancing patient care in the management of persistent AF patients.

Methods

Study population

To be eligible for this study, patients had to have persistent atrial fibrillation, and regardless of prior ablation therapy, their most recent electrical cardioversion had to have been unsuccessful. This retrospective cohort study enrolled 125 consecutive patients diagnosed with persistent AF who underwent unsuccessful ECV at a single high-volume tertiary hospital between November 2017 and August 2023. Patients were required to be aged > 18 years to participate in the study. Based on the subsequent rhythm control strategy, patients were divided into two groups: non-ablation therapy (n = 64) and ablation therapy (n = 61) (Fig. 1). Due to the retrospective nature of the study, The Institutional Review Board of Samsung Medical Center, Seoul, Korea (IRB number 2024-02-017-001) waived the need of obtaining informed consent. The study was conducted in accordance with the principles of the Declaration of Helsinki.

Study protocol

All patients received oral anticoagulation therapy for at least 3 weeks prior to the ECV^9,10. Baseline evaluations, including complete blood count, serum creatinine, and transthoracic and transesophageal echocardiograms were performed for all patients. Sedation was achieved using midazolam and propofol, with flumazenil used in most cases post-ECV. Electrocardiograms and Holter monitoring reports from November 2017 to June 2024 were obtained for all patients.

Electrical cardioversion procedure

All ECV procedures were conducted in the electrophysiology laboratory. A stepwise ECV approach was employed, consisting of three steps: Step 1—synchronized biphasic shock of 150 J using self-adhesive gel electrodes in the antero-apical location; Step 2—synchronized biphasic shock of 200 J using the same technique as in Step 1; Step 3—synchronized biphasic shock of 200 J using self-adhesive gel electrodes in the anteroposterior location.

Definitions

An unsuccessful ECV was defined as either the failure to immediately restore sinus rhythm after up to three external ECV attempts or the recurrence of atrial fibrillation during the observation period while the patient was recovering from sedation after initially restoring sinus rhythm. The patients showed no recovery of sinus rhythm at first ECV trial experienced unsuccessful ECV in most cases. The ablation therapy group included patients who underwent one or more of the following treatments: radiofrequency catheter ablation, cryoablation, and totally thoracoscopic surgical ablation. AF duration was defined as the duration from the date of AF diagnosis to the date of ECV therapy.

Ablation methods

Radiofrequency and cryoballoon ablations were performed with sedation or general anesthesia. In cases of general anesthesia, we measured esophageal temperature. Femoral vein access was obtained primarily using the Seldinger technique without ultrasound guidance. Trans-septal access and catheter navigation were performed using an SL-1 sheath (Fast-Cath guiding introducer, St. Jude Medical, Inc., St. Paul, MN, USA). 3D catheter orientation and mapping of the left atrium and the pulmonary veins were performed with a CARTO (Biosense Webster, Diamond Bar, CA, USA) or EnSite Precision (Abbott, St. Paul, MN, USA) electroanatomic mapping system. Cryoablation was performed with an Arctic Front Advance Pro (Medtronic, Minneapolis, MN, USA) or PolarX (Boston Scientific, St. Paul, MN, USA) cryoballoon system. Additional linear ablation was performed according to operator preference.

Totally thoracoscopic ablation refers to ablation with video-assisted thoracoscopy without a robotic surgery system like the Da Vinci system (Intuitive Surgical, Sunnyvale, CA, USA) or cardiopulmonary bypass. Pulmonary vein isolation was performed by clamping with bipolar radiofrequency energy. A linear pen device was used for creation of superior and inferior lines connecting the two PV isolation lines. Ablations of ganglionated plexus and ligament of Marshall were performed. An endoscopic stapling device was used for removal of the left atrial appendage. In our center, the same experienced cardiothoracic surgeon performed all surgical procedures.

Statistical analysis

All continuous variables were checked for normality using Kolmogorov–Smirnov test. Continuous variables with normal distributions were expressed as mean ± standard deviation, while those with non-normal distributions were expressed as median and interquartile range, respectively. Categorical variables were presented as number and percentage. Continuous variables with a normal distribution were compared using the Student’s t-test, while those with a non-normal distribution were compared using the Mann–Whitney U test. Nominal variables were compared using the Chi-square test. Cumulative AF-free survival was analyzed using the Kaplan–Meier product method. Patients were censored at the time of documented AF recurrence or latest follow-up electrocardiogram (or Holter monitoring). Hazard ratios for AF recurrence were estimated using Cox regression analysis, and 95% confidence intervals were reported. Univariate and multivariate Cox regression analysis were performed to assess the predictive value of the variables for AF recurrence risk. P-values ≤ 0.05 and ≤ 0.10 were considered statistically significant for outcome and predictor analysis, respectively. Statistical analyses were conducted using SPSS Statistics for Windows, version 20.0.0 (SPSS Inc., Chicago, IL, USA).

Electronic supplementary material

Below is the link to the electronic supplementary material.

Author contributions

Conceptualization, HJL, YKO; methodology, HJL, KMP, YKO; data curation, HJL, SHL, JYK, SJP, KMP; formal analysis, HJL, SHL, JWK; writing, HJL, SHL, JWK; review and editing, JYK, SJP, KMP; supervision, SJP, KMP, YKO. All authors have read and agreed to the manuscript.

Data availability

The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Competing interests

The authors declare no competing interests.