Outcomes after out-of-hospital cardiac arrest are gen-erally believed to be poor.1,2 Nevertheless, several
studies have shown significant improvements in survival rates after out-of-hospital cardiac arrest during the past 2 decades.3–5 In Denmark, multiple national initiatives in the past decade were associated with an ≈3-fold increase in survival at day 30 and 1 year after out-of-hospital car-diac arrest.3 These markedly improved survival rates call on further research on functional and neurologic outcomes in cardiac arrest survivors.
Clinical Perspective on p 1690
Studies on long-term neurologic outcome, including details of functional recovery, are requested in a consensus statement from the American Heart Association, which also focuses on the current scarcity and incomplete follow-up of such data.6 This document also notes that no single definition of neuro-logic outcome is available and that outcome assessment is a multimodal approach.6 Return to work with maintenance of salary indicates a capability to continue to function in society in the same manner as before an event but is sparsely studied after out-of-hospital cardiac arrest.7–14 In a study of 133 coma-tose out-of-hospital cardiac arrest survivors who were treated with therapeutic hypothermia, 74% of patients with favorable
Background—Data on long-term function of out-of-hospital cardiac arrest survivors are sparse. We examined return to work as a proxy of preserved function without major neurologic deficits in survivors.
Methods and Results—In Denmark, out-of-hospital cardiac arrests have been systematically reported to the Danish Cardiac Arrest Register since 2001. During 2001–2011, we identified 4354 patients employed before arrest among 12 332 working-age patients (18–65 years), of whom 796 survived to day 30. Among 796 survivors (median age, 53 years [quartile 1–3, 46–59 years]; 81.5% men), 610 (76.6%) returned to work in a median time of 4 months [quartile 1–3, 1–19 months], with a median time of 3 years spent back at work. A total of 74.6% (N=455) remained employed without using sick leave during the first 6 months after returning to work. This latter proportion of survivors returning to work increased over time (66.1% in 2001–2005 versus 78.1% in 2006–2011; P=0.002). In multivariable Cox regression analysis, factors associated with return to work with ≥6 months of sustainable employment were as follows: (1) arrest during 2006–2011 versus 2001–2005, hazard ratio (HR), 1.38 (95% CI, 1.05–1.82); (2) male sex, HR, 1.48 (95% CI, 1.06–2.07); (3) age of 18 to 49 versus 50 to 65 years, HR, 1.32 (95% CI, 1.02–1.68); (4) bystander-witnessed arrest, HR, 1.79 (95% CI, 1.17–2.76); and (5) bystander cardiopulmonary resuscitation, HR, 1.38 (95% CI, 1.02–1.87).
Conclusions—Of 30-day survivors employed before arrest, 76.6% returned to work. The percentage of survivors returning to work increased significantly, along with improved survival during 2001–2011, suggesting an increase in the proportion of survivors with preserved function over time. (Circulation. 2015;131:1682-1690. DOI: 10.1161/CIRCULATIONAHA.114.011366.)
Circulation is available at http://circ.ahajournals.org DOI: 10.1161/CIRCULATIONAHA.114.011366
Continuing medical education (CME) credit is available for this article. Go to http://cme.ahajournals.org to take the quiz.Received May 24, 2014; accepted March 5, 2015.From Department of Anesthesiology and Intensive Care Medicine, Cardiovascular Research Center (K.K., B.S.R.), and Departments of Social Medicine (K.F.)
and Cardiology (S.E.J.), Aalborg University Hospital, Denmark; Clinical Institute of Medicine, Aarhus University, Aalborg, Denmark (K.K., E.F.C., B.S.R.); Department of Cardiology, Copenhagen University Hospital, Gentofte, Denmark (M.W., S.R., G.G.); Department of Health, Science, and Technology, Aalborg University, Denmark (R.N.M., K.F., S.E.J., C.T-P.); Prehospital Emergency Medical Services, Regions of Denmark: The Capital (F.K.L.), Zealand (O.M.H.), South Denmark (T.L.-J.), Central (E.F.C.), and North Denmark (P.A.H.); Department of Cardiology, Rigshospitalet, Copenhagen University Hospital, Denmark (L.K.); National Institute of Public Health, University of Southern Denmark, Copenhagen, Denmark (G.G.); and Danish Heart Foundation, Copenhagen, Denmark (G.G.).
The online-only Data Supplement is available with this article at http://circ.ahajournals.org/lookup/suppl/doi:10.1161/CIRCULATIONAHA. 114.011366/-/DC1.
Correspondence to Kristian Kragholm, MD, Clinical Institute of Medicine, Aarhus University, and Department of Anesthesiology and Intensive Care Medicine, Cardiovascular Research Center, Aalborg University Hospital, Sdr Skovvej 15, 9000 Aalborg, Denmark. E-mail: [email protected]
Return to Work in Out-of-Hospital Cardiac Arrest SurvivorsA Nationwide Register-Based Follow-Up Study
Freddy Knudsen Lippert, MD; Erika Frischknecht Christensen, MD; Poul Anders Hansen, MD; Torsten Lang-Jensen, MD; Ole Mazur Hendriksen, MD; Lars Kober, MD, DSc;
Gunnar Gislason, MD, PhD; Christian Torp-Pedersen, MD, DSc; Bodil Steen Rasmussen, MD, PhD
Resuscitation Science
by guest on May 11, 2015http://circ.ahajournals.org/Downloaded from by guest on May 11, 2015http://circ.ahajournals.org/Downloaded from by guest on May 11, 2015http://circ.ahajournals.org/Downloaded from by guest on May 11, 2015http://circ.ahajournals.org/Downloaded from by guest on May 11, 2015http://circ.ahajournals.org/Downloaded from by guest on May 11, 2015http://circ.ahajournals.org/Downloaded from by guest on May 11, 2015http://circ.ahajournals.org/Downloaded from by guest on May 11, 2015http://circ.ahajournals.org/Downloaded from by guest on May 11, 2015http://circ.ahajournals.org/Downloaded from by guest on May 11, 2015http://circ.ahajournals.org/Downloaded from by guest on May 11, 2015http://circ.ahajournals.org/Downloaded from
discharge cerebral performance category scores had returned to work at 1-year follow-up.7 In patients surviving stroke, sur-vivors with favorable discharge modified Rankin Scale neuro-logic outcome scores returned to work faster and to a greater extent than patients with less favorable discharge scores.15–17 In addition, among stroke survivors with favorable modified Rankin scores, differences in higher cortical and cognitive function were present and had further significant impacts on return-to-work rates, supporting return to work as a proxy of preserved function without major neurologic deficits.15
Anoxic brain injury represents a severe damage to the brain caused by insufficient oxygen supply during cardiac arrest, which often leads to limitations in cognitive function and activities of daily living.2,18 Despite the 3-fold increase in the survival rates after out-of-hospital cardiac arrest in Denmark, the extent to which survivors return to their prearrest func-tion remains unknown. Therefore, we examined return to work, maintenance of work, and income in a nationwide study including 30-day survivors of out-of-hospital cardiac arrest who were employed up until the incident of cardiac arrest. We further evaluated first-time incidence of anoxic brain damage in the time frame from index hospitalization to 30 days after discharge. We hypothesized that the significant increase in sur-vival rates during the past decade was followed by a concomi-tant increase in the proportion of patients returning to work.
MethodsStudy SettingThis study was conducted from June 1, 2001, to December 31, 2011, in Denmark, which covers 43,000 km2, including rural, suburban, and urban areas.
The emergency medical service consists of basic life support ambulances staffed with ambulance technicians or paramedics and mobile emergency care units staffed with paramedics, nurses, or anesthesiologists. The mobile emergency care units are dispatched as rendezvous with basic life support ambulances. Throughout the study period, treatment has been given according to the latest guidelines.19
Study PopulationThe national healthcare system of Denmark, including the emer-gency medical service system, is tax financed, assuring free access to emergency medical service assistance in Denmark and meaning that all Danish citizens in the study period contribute to form the source population. During the study period, the source population increased from 5 355 000 (3 443 533 between 18 and 65 years of age) to 5 581 000 (3 486 982 between 18 and 65 years of age).20
Patients experiencing out-of-hospital cardiac arrest are included in the national Danish Cardiac Arrest Registry when resuscitation efforts are initiated either by bystanders (with activation of the emer-gency medical service) or by emergency medical service personnel. Patients with obvious late signs of death (ie, rigor mortis), where resuscitative efforts are not initiated, are not included in the defini-tion of out-of-hospital cardiac arrest. Since the emergency medical service system is activated for all emergencies concerning out-of-hospital cardiac arrest in Denmark and the emergency medical ser-vice personnel through contractual commitments are required to fill in a short case report form for the Danish Cardiac Arrest Registry for every out-of-hospital cardiac arrest, case ascertainment is close to complete. An average incidence rate of 57.4 per 100 000 persons per year in the years 2001–2011 was reported previously for the Danish Cardiac Arrest Registry and is comparable to both European and American contemporary data.21–24 Because we studied return to work, only 30-day survivors of working age between 18 and 65 years, who were working up until arrest, were relevant for inclusion.
Study DesignThis nationwide study uses a register-based follow-up design. From the Danish Cardiac Arrest Registry, we included information on date, time, and year of arrest; the site of arrest (at home versus not at home); whether the collapse was unwitnessed or witnessed either by a bystander or by the emergency medical service; whether the bystander initiated cardiopulmonary resuscitation or defibrillated the patient; the time interval from the emergency call to rhythm analysis by the emergency medical service; whether the first recorded heart rhythm was shockable (pulseless ventricular tachycardia or ventricu-lar fibrillation) or nonshockable (asystole or pulseless electric activ-ity); and whether return of spontaneous circulation was achieved at the time of hospital arrival.
We used the Danish Civil Personal Registration Number unique to each Danish citizen to link information from the Danish Cardiac Arrest Register with other national registries.25 Data on age, sex, and status of living alone were retrieved from the Civil Personal Registration Registry.25 The presumed cause of arrest was assessed from the Danish National Patient Register and the Danish Cause of Death Register, using information from death certificates and dis-charge diagnoses. Patients with cardiac disease, unknown disease, or unexpected collapse were categorized as having a presumed cardiac cause of arrest. Death certificates and discharge diagnoses with other medical disorders (apart from conditions mentioned above) were defined as noncardiac causes.
Employment status was obtained from a registry administered by the Danish National Labor Market Authority and available to the study on a weekly basis for baseline and follow-up assessment until February 24, 2013.26 In accordance with the definition of self-supporting, previously validated with a positive predictive value of 98.2%, patients who were not on any social benefits or on State Education Fund grants, maternity leave pay, or leave-of-absence schemes were classified as working.26 Accordingly, patients receiv-ing unemployment benefits or sickness leave pay or who were on early retirement payment were defined as being on social benefits. We assessed employment status in a 5-week span before cardiac arrest to classify patients as either working or receiving social ben-efits. Patients on social benefits at baseline were excluded, and return to work was evaluated in survivors working at baseline only. This 5-week evaluation of employment status was used both at baseline and 1-year follow-up to reduce misclassification; that is, if sick leaves were <3 weeks during the 5-week span, patients were classified as working.27
Other than evaluating employment status at baseline and 1-year follow-up, we also calculated time to return to work, counting from the day of entry (30 days/4 weeks after arrest) to the first week of employment. For patients returning to work, we further evaluated the duration of maintenance of work. It has been shown previously that a significant proportion of patients with coronary heart disease under-going percutaneous intervention treatment with return to work have recurrent sick leaves and, thus, return to work alone may not indicate full recovery in all cases.27 Therefore, we further assessed definitions of maintenance of work in which we allowed for 4 to 12 weeks of work absences after return to work. Further in line with this study, we defined return to work without any sick leave relapse during the first 6 months of employment as our end point for modeling. Thus, for modeling, survivors were followed from day 30 until the end of February 2013 and were censored for death or for not meeting this end point definition during follow-up. To examine whether patients who met this end point definition were able to maintain the same level of income as before arrest, we linked employment data with annual income information, comparing the income level in the year before arrest with the income level in the year after arrest in patients.
Information on education and household and personal income was available on a yearly basis and was attained from the Statistics of Denmark, and we indexed income for 2010.28,29 We further linked employment data with industry affiliation information available to the study from the Danish National Labor Market Authority registry on a monthly basis since 2008 to distinguish income based on salary from income based on insurance or other private means of compensation for loss of ability to work.
by guest on May 11, 2015http://circ.ahajournals.org/Downloaded from
To supply the employment outcomes, we assessed the onset of anoxic brain damage in the time frame from index hospitalization to 30 days postdischarge, using information on discharge diagnoses from the Danish National Patient Register.30 To evaluate first-incident cases only, we intended to exclude patients with anoxic brain dam-age before arrest; however, no patients were found to have this pre-existing condition.
Data on comorbidity, length of hospital stay, and medication use before arrest were retrieved from the Danish National Patient Register and the Danish National Prescription Registry.30,31 To obtain the Charlson Comorbidity Index Score, discharge diagnoses and certain medication use were assessed up to 10 years before cardiac arrest.32
OutcomesPrimary outcome was return to work. Secondary outcomes were maintenance of work and salary, as well as new onset of anoxic brain damage after arrest.
Statistical AnalysesWe summarized categorical variables using percentages and frequen-cies and continuous variables using medians and 1st to 3rd quartiles (Q1–Q3) by year of arrest (see Table 1). Cumulative incidences for time to return to work and maintenance of work were depicted. For univariable Cox regression analyses, observations with missing data were excluded in the calculations (see Table 2). For multivariable Cox regression analyses, missing data were imputed using multiple impu-tation methods (Multivariate Imputation in Chained Equations).33 Applying these methods, 10 imputed data sets (complete data sets with observed and imputed values) were constructed using informa-tion from all of the patient- and cardiac arrest–related variables in Table 1. Estimates were then calculated from the imputed data sets and combined using multiply imputed pooled analysis.33 We further compared results derived from the pooled analyses of imputed data sets to results derived from analyses of sensitivity analysis data sets constructed by replacing missing values in the original data set with the worst, best, or neutral values (worst-best case scenarios). Before imputation, missing data patterns and missing data mechanisms were investigated, and we found no indication of informative missing.
Covariates for our multivariable model were selected before analy-sis, and we included those we considered most relevant based on cur-rent knowledge and literature (sex, age, year of arrest, status of living alone, baseline income and educational levels, Charlson comorbidity score, arrest witnessed by bystander or emergency medical service or not witnessed, bystander cardiopulmonary resuscitation, presumed cardiac cause of arrest, and length of stay in hospital).3–5,34 To explore the role of bystander cardiopulmonary resuscitation, we excluded emergency medical service–witnessed arrest cases from our model. No significant collinearity was found and no further selection was
performed. The assumption of proportional hazards was checked by plotting Schoenfeld residuals by time for each covariate, and no time trends were seen. A paired t test was used to test whether sur-vivors who returned to work maintained the same annual individual and household incomes in the year after arrest compared with the year before arrest. A 2-sided P value <0.05 was considered statisti-cally significant. Data management and statistical analyses were per-formed using SAS version 9.2 (SAS Institute Inc, Cary, NC) and R version 3.0.2 (R Development Core Team).
EthicsThe study was approved by the Danish Data Protection Agency (2007-58-0015, GEH- 2014–017, I-Suite-nr 02735). In Denmark, ethical approval is not required for register-based studies.
ResultsPatientsFrom June 1, 2001, to December 31, 2011, 12 332 patients who experienced out-of-hospital cardiac arrest in the Danish Cardiac Arrest Registry were of working age (18–65 years). Of 4354 patients working at baseline, 796 patients survived to day 30 and formed the study population (see Figure 1). Patient characteristics by year of arrest are shown in Table 1.
Time to Return to Work and Maintenance of WorkThe cumulative incidence of patients returning to work (N=610) and patients dying (N=34) during follow-up until the end of February 2013 is shown in Figure 2; accordingly, 152 patients did not return to work during follow-up. A total of 76.6% of the survivors (N=610) returned to work in a median time of 4 months (Q1–Q3, 1–19 months).
Maintenance of work was assessed for the 610 survivors who returned to work (see Figure 3). These survivors remained employed for a median time of 3 years, when no work absences after return to work were tolerated. When accepting sickness leaves ≤4 and 12 weeks after return to work as ongo-ing employment, median time to withdrawal from work then increased to 5 years and 11 years, respectively (see Figure 3).
Factors Associated With Return to WorkAmong the 610 patients returning to work, the proportion of patients maintaining work without recurrent work absences
Table 1. Baseline, Prehospital, and In-Hospital Characteristics and Outcomes by Year of Arrest
Parameter Statistic
2001–2005 (N=237)
2006–2011 (N=559) All Years (N=796)
n Data n Data n DataBaseline variablesAge Median, Q1 to Q3 53 46–59 53 45–59 53 46–59Sex Men, n (%) 195 82.3 454 81.2 649 81.7
Women, n (%) 42 17.7 105 18.8 147 18.3Living alone Yes, n (%) 49 20.7 138 24.7 187 22.7
No, n (%) 188 79.3 419 75.0 607 77.1Missing 0 2 0.4 2 0.2
Household income* < 1st quartile, n (%) 53 22.4 144 25.8 197 24.11st to 3rd quartiles, n (%) 129 54.4 263 47.0 392 50.7>3rd quartile, n (%) 55 23.2 141 25.2 196 24.2
(Continued)
by guest on May 11, 2015http://circ.ahajournals.org/Downloaded from
Hospital length of stay, d<5 n (%) 69 29.1 175 31.3 244 30.25–12 n (%) 106 44.7 238 42.6 344 43.7>12 n (%) 58 24.5 136 24.3 194 24.4
Missing 4 1.7 10 1.8 14 1.7Outcomes1-y RTW n (%) 125 52.7 340 60.8 465 56.8Any RTW during follow-up n (%) 177 74.7 433 77.5 610 76.1RTW without work absences in first 6 months of work§
n (%) 117 66.1 338 78.1 455 74.6
Death at 1-y follow-up n (%) 10 4.2 12 2.1 22 3.2Death at any time during follow-up
n (%) 16 6.8 18 3.2 34 5.0
AED indicates automatic external defibrillator; CPR, cardiopulmonary resuscitation; EMS, emergency medical service; Q1–Q3, 1st to 3rd quartile; ROSC, return of spontaneous circulation; RTW, return to work; and VT/VF, ventricular tachycardia/fibrillation.
*Household income is indexed for the year 2010 and grouped according to quartiles: group 1 < $79 679, group 2 $79 679-$149 629 (Q1-Q3), and group 3 > $149 629.
†Educational level is as follows: group 1 includes educational level of the International Standard Classification of Education (ISCED) 0–2 (preprimary, primary, and lower secondary education); group 2 includes ISCED 3+5 (ISCED 4 does not exist in Denmark; high school or vocational education and short-cycle tertiary education); and group 3 includes ISCED 6–8 (median-length tertiary education, bachelor, master, and PhD degrees).
‡Data include the time interval from recognition of arrest (time of emergency call) to rhythm analysis by emergency medical service.§Data are among the 610 patients who returned to work.
Table 1. Continued
Parameter Statistic
2001–2005 (N=237)
2006–2011 (N=559) All Years (N=796)
n Data n Data n Data
by guest on May 11, 2015http://circ.ahajournals.org/Downloaded from
during the first 6 months of employment (N=455) increased over time (66.1% [N=117] in 2001–2005 versus 78.1% [N=338] in 2006–2011; P=0.002). Arrest in 2006–2011 versus
2001–2005, male sex, high educational and income levels, witnessed arrests, and bystander cardiopulmonary resuscita-tion were significantly associated with this sustainable return to work definition in univariable Cox regression analyses (see Table 2). In multivariable Cox regression analysis, arrest in 2006–2011 versus 2001–2005, age between 18 and 49 ver-sus 50 and 65 years, male sex, high household income and high educational levels, and bystander- and emergency medi-cal service–witnessed arrests were significantly associated with sustainable return to work (see Figure 4). Excluding emergency medical service–witnessed arrest cases from our model introduced bystander cardiopulmonary resuscitation as a significant factor associated with sustainable return to work (see Figure 5).
We performed a number of sensitivity analyses. Findings from analyses shown in Figures 4 and 5 using imputed data were not substantially different from analyses, in which cases with missing data were excluded (see Tables I and II in the online-only Data Supplement). Because patients were included until December 31, 2011, and follow-up time ended on February 24, 2013, we may have underestimated the num-ber of survivors returning to work in the end of our study period. Therefore, because all of the patients had at least 1 year of follow-up time, we also assessed return to work in the prespecified 5-week span around 1-year follow-up, in which 465 patients were working and 331 patients were not, of which 22 had died (see Table 1). No substantial differences were seen between findings from univariable and multivari-able logistic regression analyses of these patients and uni-variable and multivariable Cox regression analyses shown in Table 2 and Figures 4 and 5 (see Tables III and V and Figures I and II in the online-only Data Supplement).
Maintenance of SalaryData on income was available to the study on a yearly basis. Postarrest individual and household income levels were com-parable with the corresponding prearrest amounts among the 455 survivors returning to work without sick leave relapse in the first 6 months of employment. Using paired t tests with income indexed for the year 2010, we compared median indi-vidual income 1 year before arrest with 1 year after arrest, $86 347 (Q1–Q3: $66 818–$110 487) versus $85 647 (Q1–Q3: $66 099–$110 936; P=0.15), and median household income 1 year before arrest versus 1 year after arrest, $122 939 (Q1–Q3: $86 381–$158 789) and $125 713 (Q1–Q3: $90 574–$161 919; P=0.067).
Information regarding industry affiliation was available to the study on a monthly basis from the year 2008 onward. In a subanalysis including patients with return to work and data on industry affiliation from 2008 to 2011 (315 of 610 patients), 91.7% of these patients (N=289) had industry affiliations, indicative of income from salary.
Anoxic Brain DamageA significant decline in new onset of anoxic brain damage between 2001 and 2011 was seen (P=0.015), as 31 (13.1%) of 237 survivors in 2001–2005 compared with 42 (7.5%) of 559 survivors in 2006–2011 were diagnosed with new onset of anoxic brain damage.
Table 2. Univariable Unadjusted Cox Regression Analyses of Return to Work
Parameter Crude HR 95% Low 95% High P Value
Male sex 1.36 1.05 1.75 0.018
Age of 18–49 versus 50–65 y
1.17 0.97 1.41 0.11
Not living alone 1.14 0.91 1.42 0.27
Household income group 2* 1.09 0.86 1.39 0.47
Household income group 3* 1.86 1.44 2.41 <0.001
Education group 2† 1.40 1.12 1.75 0.003
Education group 3† 1.91 1.47 2.48 <0.001
Charlson score 0 vs >0 1.24 0.95 1.62 0.12
Arrest witnessed by bystander
1.80 1.19 2.70 0.005
Arrest witnessed by EMS 2.20 1.43 3.38 <0.001
Bystander CPR‡ 1.60 1.20 2.13 0.001
Presumed cardiac cause of arrest
1.55 0.95 2.51 0.077
Arrest in years 2006–2011 versus 2001–2005
1.32 1.07 1.63 0.010
Length of stay in hospital (5–12 d)
0.85 0.69 1.06 0.14
Length of stay in hospital (>12 d)
0.75 0.58 0.96 0.022
Table includes return to work without work absence leaves during the first 6 months of employment. CPR indicates cardiopulmonary resuscitation; EMS, emergency medical service; and HR, hazard ratio.
*Household income is indexed for the year 2010 and median is $115 447 (quartile 1–3, $79 679–$149 629).
†Educational level is as follows: group 1 includes educational level of the International Standard Classification of Education (ISCED) 0–2 (preprimary, primary, and lower secondary education); group 2 includes ISCED 3+5 (ISCED 4 does not exist in Denmark; high school or vocational education and short-cycle tertiary education); and group 3 includes ISCED 6–8 (median-length tertiary education, bachelor, master, and PhD degrees).
‡Data include bystander CPR in non–EMS-witnessed cases.
36,987 patients with out-of-hospital cardiac arrest in whom resuscitation was attempted
20,551 patients of age below 18 years or above 65 years
12,332 patients of working age between 18 and 65 years of age
3,953 (10.7%) were excluded because of invalid or missing civil registration number89 (0.2%) were excluded due to arrest occurring more than once62 (0.2%) were excluded due to missing data concerning hospital admission
3,558 patients not alive at day 30
32,883 patients eligible for further assessment
4,354 patients working at baseline
418 patients without employment data7,560 patients not working at baseline
Final study population796 patients working at baselineand alive at day 30
Figure 1. Patient selection, from the Danish Cardiac Arrest Registry, 2001–2011. Data show a flowchart of the selection process for the study population.
by guest on May 11, 2015http://circ.ahajournals.org/Downloaded from
DiscussionThis nationwide study demonstrates that the tripling in sur-vival rates associated with multiple national initiatives to enhance survival in Denmark in the past decade was accom-panied by a substantially increased proportion of patients returning to work with maintenance of salary. More than 75% of the studied patients returned to work, and these patients maintained their work status for a long time onward. In tan-dem with the increase in the rate of return to work, a signifi-cant decline in the proportions of patients with onset of anoxic brain damage after arrest was observed, supporting return to work as a proxy of preserved capability to work without major neurologic deficits.
In previous studies of return to work after out-of-hospital cardiac arrest, return to work rates varied between 15% and 79%.7–14 In line with studies of favorable return-to-work rates, we report an overall return to work rate of 76.6% and a 1-year status of 58.4%. Our median time of 4 months from day 30 after arrest to return to work is also in line with previous median return-to-work times of ≈6 months after out-of-hos-pital cardiac arrest.11,12
Previous studies on return to work after out-of-hospital cardiac arrest were limited to small subgroups of patients,
with descriptive statistics or univariable analyses only.7–14 Our study is the first of its kind to examine factors associ-ated with return to work in multivariable-adjusted analyses. The impacts of bystander- and emergency medical service–witnessed arrest on return to work likely reflect the ben-eficial effect of early resuscitative efforts, as seen in other studies.3,4 Benefit from early resuscitative efforts is further supported by the association of bystander cardiopulmonary resuscitation and return to work, when emergency medical service–witnessed arrests were excluded. During the past decade, multiple nationwide initiatives taken in Denmark were associated with a doubling in bystander cardiopulmo-nary rates and a tripling in survival rates after out-of-hospital cardiac arrest.3 These initiatives include widespread cardio-pulmonary resuscitation training, along with implementation of resuscitation guideline changes in 2005 onward, stress-ing the importance of chest compressions in resuscitative efforts.3,19 Furthermore, mobile emergency care units staffed with anesthesiologists or paramedics increased during the study period, and healthcare professionals at dispatch centers began providing telephone-assisted instructions to bystand-ers from 2009. Throughout the study period, there was a substantial increase in the number of automated external
Figure 3. Maintenance of work after return to work after out-of-hospital cardiac arrest. Data show the duration of maintenance of work counting from the week of return to work (for the 610 survivors who returned to work) and followed until February 24, 2013. Vertical axis, cumulative incidence (%). Horizontal axis, time (years). The color black depicts a median rate of 3 years of time to withdrawal from work when no short sickness leaves were tolerated as ongoing employment; the color red depicts a median rate of 5 years to withdrawal, when sickness leaves of ≤4 weeks after return to work were not regarded as withdrawal from work; and the color blue depicts a median of 11 years to withdrawal, when sickness leaves ≤12 weeks after return to work were not categorized as withdrawal from work.
Figure 2. Time to return to work after out-of-hospital cardiac arrest. Data show the time to return to work counting from day 30 after cardiac arrest and followed until February 24, 2013. Vertical axis, cumulative incidence (%). Horizontal axis, time (years). Median time to return to work, with 1st to 3rd quartiles (Q1–Q3), 4 months (Q1–Q3, 1–19 months).
by guest on May 11, 2015http://circ.ahajournals.org/Downloaded from
defibrillators located outside hospitals, although the use and increase in use of such defibrillators by bystanders were lim-ited in our study. Finally, intensive care including therapeutic hypothermia was implemented as standard care for comatose survivors of cardiac arrest since 2004 in Denmark.3 Although these observational data cannot conclude which factors have contributed most to the changes observed, our findings sug-gest that the sum of national initiatives taken to improve car-diac arrest management not only improved survival but also increased return-to-work rates over time. Lastly, the asso-ciations of high educational and income levels and return to work may reflect easier return to white- versus blue-collar jobs, as seen in stroke survivors, with potential implications for return-to-work intervention programs, but deserves fur-ther study.17,34
LimitationsUsing nationwide registries, we minimized the risk of selec-tion bias attributed to geographic differences in patient- and cardiac arrest–related characteristics. Furthermore, the unique Civil Personal Registration Number given to each Danish citizen allowing linkage between medical data and follow-up registries resulted in an extremely low lost-to-follow-up rate. Nonetheless, the study had some limitations.
First, because of the observational study design, results are associations and may not be causal.
Second, a number of patients had missing data. However, results of multivariable analyses with missing data excluded did not substantially differ from results based on pooled anal-yses of imputed data sets. Thus, missing data were unlikely to have influenced our main conclusions.
Figure 4. Multivariable Cox regression analysis of return to work. Data show return to work without work absence leaves during the first 6 months of employment. Household income was indexed for year 2010 and categorized according to quartiles, with group 1 as reference: group 1, <$79 679; group 2, 1st to 3rd quartile (Q1-Q3) $79 679 to $149 629; group 3, >$149 629. Educational level was divided into 3 groups, with group 1 as reference: group 1/International Standard Classification of Education (ISCED) 0 to 2 (preprimary, primary, and lower secondary education); group 2/ISCED 3+5 (high school or vocational education and short-cycle tertiary education; ISCED 4 does not exist in Denmark); and group 3/ISCED 6 to 8 (medium-length tertiary education, bachelor, master, and PhD degrees). Length of stay in hospital is divided into 3 groups, with 0 to 4 days as reference. CPR indicates cardiopulmonary resuscitation; EMS, emergency medical service.
Figure 5. Multivariable Cox regression analysis of return to work (EMS-witnessed arrests excluded). Data show return to work without work absence leaves during the first 6 months of employment. Household income was indexed for year 2010 and categorized according to quartiles, with group 1 as reference: group 1, <$79 679; group 2, quartile 1 to 3 (Q1-Q3), $79 679 to $149 629; group 3, >$149 629. Educational level was divided into 3 groups, with group 1 as reference: group 1/International Standard Classification of Education (ISCED) 0 to 2 (preprimary, primary, and lower secondary education); group 2/ISCED 3+5 (high school or vocational education and short-cycle tertiary education, ISCED 4 does not exist in Denmark); and group 3/ISCED 6 to 8 (medium-length tertiary education, bachelor, master, and PhD degrees). Length of stay in hospital is divided into 3 groups, with 0 to 4 days as reference. EMS indicates emergency-witnessed arrest; CPR, cardiopulmonary resuscitation.
by guest on May 11, 2015http://circ.ahajournals.org/Downloaded from
Third, since our study outcome was return to work, only 30-day survivors of working age who were working before arrest were relevant to the study. However, the concomitant decline in anoxic brain damage found in this study was also seen in a study of all nonemergency medical service–wit-nessed out-of-hospital cardiac arrest survivors of presumed cardiac cause in Denmark during 2001–2010.3 Extrapolation of our data to account for improvement in function in all sur-vivors in Denmark during the past decade may therefore be justified but remains to be further explored.
Fourth, employment status could be subject to misclas-sification.26,27 However, the high positive predictive value of being registered as self-supporting in the employment registry, together with the applied 5-week spans to evaluate employment status at baseline and for 1-year follow-up, the sustainable return-to-work definition for modeling, and sub-sequent linkages to income and industry affiliation minimized potential misclassification. Few of the studied patients who were registered as having returned to work (8.3%) did not have industry affiliation, which either indicates a missing data problem in industry affiliation codes or that these patients did not return to work but supported themselves by worker’s com-pensation. On the contrary, we may also have underestimated true favorable outcome. It is not unlikely that a number of patients close to the public retirement age of 65 years never returned to work or retired soon after return to work because of eligibility of regular retirement on the basis of age and not necessarily poor function. Furthermore, in the beginning of the study period, the economy in Denmark was booming and, in the end, a recession had begun, with a possible effect on the number of jobs available and the number of patients returning to work. Lastly, since patients were included until December 31, 2011, and follow-up time ended on February 24, 2013, we likely underestimated the number of survivors returning to work in the end of our study period. Nevertheless, the propor-tion of patients returning to work was significantly higher late in the study period compared with early in the period.
Fifth, although its sensitivity is unknown, we cannot think of any reasons that would lead to inequality in coding and reporting of anoxic brain injury to the Danish National Patient Registry over time. Moreover, because it is a discharge diagnosis given by a neurologist, its specificity approximates 100%. Patients with minor cerebral dysfunction were potentially not referred to a neurologist and may therefore not be captured by this diag-nosis. Regardless of minor cerebral dysfunction, return to work with maintenance of salary stands as a unique favorable outcome measure indicating the capability to continue to function in soci-ety in the same manner as before an event. The decline in sur-vivors diagnosed with anoxic brain injury concurrent with the increase in return-to-work rate supports this notion.In conclusion, the tripling in the number of out-of-hospital cardiac arrest survivors during the past decade in Denmark was accompanied by a significant increase in the proportion of patients returning to work with maintenance of salary and a significant decrease in new onset of anoxic brain damage. Approximately three fourths of the studied patients returned to work with a subsequent long duration of maintenance of work. The concurrent decline in the proportion of survivors diagnosed with anoxic brain injury, along with the increase in
the return-to-work rate, supports return to work as a proxy of preserved capability to work without major neurologic deficits.
AcknowledgmentsWe extend our sincere thanks to the Danish Emergency Medical Service personnel who have completed the case report forms for the Danish Cardiac Arrest Register.
Sources of Funding This study was supported by the Danish foundation TrygFonden and the Danish Heart Foundation. The Danish Cardiac Arrest Registry is supported by TrygFonden, which has no commercial interests in the field of cardiac arrest. None of these institutions had any influ-ence on the design and conduct of the study; collection, management, analysis, and interpretation of the data; and preparation, review, or approval of the article for submission.
Disclosures Dr Kragholm is supported by the Danish Heart Foundation, the Laerdal Foundation, and the Fund of Herta Christensen. Dr Wissenberg is supported by the TrygFond Foundation, the Danish Heart Foundation, and the Health Insurance Foundation. Dr Gislason is supported by an unrestricted research scholarship from the Novo Nordisk Foundation and has received speaker honorarium from AstraZeneca, Bristol-Myers Squibb, Sanofi-Aventis, Pfizer and Bayer and research grants from Bristol-Myers Squibb/Pfizer, AstraZeneca and Bayer. Dr Kober has received payment for speaking at a sympo-sium arranged by Servier. Dr Torp-Pedersen is serving as a consultant for Cardiome, Merck, Sanofi, and Daiichi and has received grants or has grants pending from Bristol-Myers Squibb. The other authors report no conflicts.
from out-of-hospital cardiac arrest: a systematic review and meta-anal-ysis. Circ Cardiovasc Qual Outcomes. 2010;3:63–81. doi: 10.1161/CIRCOUTCOMES.109.889576.
2. Moulaert VR, Verbunt JA, van Heugten CM, Wade DT. Cognitive impair-ments in survivors of out-of-hospital cardiac arrest: a systematic review. Resuscitation. 2009;80:297–305. doi: 10.1016/j.resuscitation.2008.10.034.
3. Wissenberg M, Lippert FK, Folke F, Weeke P, Hansen CM, Christensen EF, Jans H, Hansen PA, Lang-Jensen T, Olesen JB, Lindhardsen J, Fosbol EL, Nielsen SL, Gislason GH, Kober L, Torp-Pedersen C. Association of national initiatives to improve cardiac arrest management with rates of bystander intervention and patient survival after out-of-hospital cardiac arrest. JAMA. 2013;310:1377–1384. doi: 10.1001/jama.2013.278483.
4. Hollenberg J, Herlitz J, Lindqvist J, Riva G, Bohm K, Rosenqvist M, Svensson L. Improved survival after out-of-hospital cardiac arrest is associated with an increase in proportion of emergency crew-wit-nessed cases and bystander cardiopulmonary resuscitation. Circulation. 2008;118:389–396.
5. Kitamura T, Iwami T, Kawamura T, Nitta M, Nagao K, Nonogi H, Yonemoto N, Kimura T; Japanese Circulation Society Resuscitation Science Study Group. Nationwide improvements in survival from out-of-hospital cardiac arrest in Japan. Circulation. 2012;126:2834–2843. doi: 10.1161/CIRCULATIONAHA.112.109496.
6. Becker LB, Aufderheide TP, Geocadin RG, Callaway CW, Lazar RM, Donnino MW, Nadkarni VM, Abella BS, Adrie C, Berg RA, Merchant RM, O’Connor RE, Meltzer DO, Holm MB, Longstreth WT, Halperin HR; American Heart Association Emergency Cardiovascular Care Committee; Council on Cardiopulmonary, Critical Care, Perioperative and Resuscitation. Primary outcomes for resuscitation science studies: a consensus statement from the American Heart Association. Circulation. 2011;124:2158–2177. doi: 10.1161/CIR.0b013e3182340239.
7. Kragholm K, Skovmoeller M, Christensen AL, Fonager K, Tilsted HH, Kirkegaard H, DE Haas I, Rasmussen BS. Employment status 1 year after out-of-hospital cardiac arrest in comatose patients treated with therapeutic hypo-thermia. Acta Anaesthesiol Scand. 2013;57:936–943. doi: 10.1111/aas.12142.
8. Fugate JE, Moore SA, Knopman DS, Claassen DO, Wijdicks EF, White RD, Rabinstein AA. Cognitive outcomes of patients undergoing
by guest on May 11, 2015http://circ.ahajournals.org/Downloaded from
therapeutic hypothermia after cardiac arrest. Neurology. 2013;81:40–45. doi: 10.1212/WNL.0b013e318297ee7e.
9. Graves JR, Herlitz J, Bång A, Axelsson A, Ekström L, Holmberg M, Lindqvist J, Sunnerhagen K, Holmberg S. Survivors of out of hospital car-diac arrest: their prognosis, longevity and functional status. Resuscitation. 1997;35:117–121.
10. Hofgren C, Lundgren-Nilsson A, Esbjörnsson E, Sunnerhagen KS. Two years after cardiac arrest; cognitive status, ADL function and living situa-tion. Brain Inj. 2008;22:972–978. doi: 10.1080/02699050802491289.
11. Bergner L, Bergner M, Hallstrom AP, Eisenberg M, Cobb LA. Health sta-tus of survivors of out-of-hospital cardiac arrest six months later. Am J Public Health. 1984;74:508–510.
12. Bergner L, Hallstrom AP, Bergner M, Eisenberg MS, Cobb LA. Health status of survivors of cardiac arrest and of myocardial infarction controls. Am J Public Health. 1985;75:1321–1323.
13. Sunnerhagen KS, Johansson O, Herlitz J, Grimby G. Life after cardiac arrest; a retrospective study. Resuscitation. 1996;31:135–140.
14. Tresch DD, Keelan MH Jr, Siegel R, Troup PJ, Bonchek LI, Olinger GN, Brooks HL. Long-term survival after prehospital sudden cardiac death. Am Heart J. 1984;108:1–5.
15. Tanaka H, Toyonaga T, Hashimoto H. Functional and occupational charac-teristics associated with very early return to work after stroke in Japan. Arch Phys Med Rehabil. 2011;92:743–748. doi: 10.1016/j.apmr.2010.12.009.
16. Quinn AC, Bhargava D, Al-Tamimi YZ, Clark MJ, Ross SA, Tennant A. Self-perceived health status following aneurysmal subarachnoid haemorrhage: a cohort study. BMJ Open. 2014;4:e003932. doi: 10.1136/bmjopen-2013-003932.
17. Peters GO, Buni SG, Oyeyemi AY, Hamzat TK. Determinants of return to work among Nigerian stroke survivors. Disabil Rehabil. 2013;35:455–459. doi: 10.3109/09638288.2012.697251.
18. Middelkamp W, Moulaert VR, Verbunt JA, van Heugten CM, Bakx WG, Wade DT. Life after survival: long-term daily life functioning and quality of life of patients with hypoxic brain injury as a result of a cardiac arrest. Clin Rehabil. 2007;21:425–431. doi: 10.1177/0269215507075307.
19. Nolan JP, Deakin CD, Soar J, Böttiger BW, Smith G; European Resuscitation Council. European Resuscitation Council guidelines for resuscitation 2005: section 4–adult advanced life support. Resuscitation. 2005;67(suppl 1):S39–S86. doi: 10.1016/j.resuscitation.2005.10.009.
20. Danmarks Statistik [Statistics of Denmark]. Hver anden borger i København er ugift [Every second citizen in Denmark is unmarried]. http://www.dst.dk/pukora/epub/Nyt/2012/NR068.pdf. Accessed March 11, 2015.
21. Wissenberg M, Hansen CM, Mortensen RM, Folke F, Lippert F, Christensen EF, Hansen PA, Lang-Jensen T, Hendriksen OM, Torp-Pedersen C. Dansk Hjertestopregister. Hjertestop udenfor hospital i Danmark. Videnskabelig Rapport 2001–2011 [Danish Cardiac Arrest Registry. Out-of-hospital cardiac arrest in Denmark. Scientific Report 2001–2011]. http://genoplivning.dk/wp-content/uploads/2013/10/Dansk-Hjertestopregister-2001–2011.pdf. Accessed March 11, 2015.
22. Nichol G, Thomas E, Callaway CW, Hedges J, Powell JL, Aufderheide TP, Rea T, Lowe R, Brown T, Dreyer J, Davis D, Idris A, Stiell I; Resuscitation Outcomes Consortium Investigators. Regional variation in out-of-hospital cardiac arrest incidence and outcome. JAMA. 2008;300:1423–1431. doi: 10.1001/jama.300.12.1423.
23. McNally B, Robb R, Mehta M, Vellano K, Valderrama AL, Yoon PW, Sasson C, Crouch A, Perez AB, Merritt R, Kellermann A. Out-of-hospital cardiac arrest surveillance: Cardiac Arrest Registry to Enhance Survival (CARES), United States, October 1, 2005-December 31, 2010. MMWR Surveill Summ. 2011;60:1–19.
24. Atwood C, Eisenberg MS, Herlitz J, Rea TD. Incidence of EMS-treated out-of-hospital cardiac arrest in Europe. Resuscitation. 2005;67:75–80. doi: 10.1016/j.resuscitation.2005.03.021.
25. Pedersen CB. The Danish Civil Registration System. Scand J Public Health. 2011;39(7 suppl):22–25. doi: 10.1177/1403494810387965.
26. Hjollund NH, Larsen FB, Andersen JH. Register-based follow-up of social benefits and other transfer payments: accuracy and degree of completeness in a Danish interdepartmental administrative database compared with a population-based survey. Scand J Public Health. 2007;35:497–502. doi: 10.1080/14034940701271882.
27. Biering K, Hjøllund NH, Lund T. Methods in measuring return to work: a comparison of measures of return to work following treatment of coro-nary heart disease. J Occup Rehabil. 2013;23:400–405. doi: 10.1007/s10926-012-9405-x.
28. Baadsgaard M, Quitzau J. Danish registers on personal income and trans-fer payments. Scand J Public Health. 2011;39(7 suppl):103–105. doi: 10.1177/1403494811405098.
30. Lynge E, Sandegaard JL, Rebolj M. The Danish National Patient Register. Scand J Public Health. 2011;39(7 suppl):30–33. doi: 10.1177/1403494811401482.
31. Kildemoes HW, Sørensen HT, Hallas J. The Danish National Prescription Registry. Scand J Public Health. 2011;39(7 suppl):38–41. doi: 10.1177/1403494810394717.
32. Thygesen SK, Christiansen CF, Christensen S, Lash TL, Sørensen HT. The predictive value of ICD-10 diagnostic coding used to assess Charlson comorbidity index conditions in the population-based Danish National Registry of Patients. BMC Med Res Methodol. 2011;11:83. doi: 10.1186/1471-2288-11-83.
33. Ambler G, Omar RZ, Royston P. A comparison of imputation tech-niques for handling missing predictor values in a risk model with a binary outcome. Stat Methods Med Res. 2007;16:277–298. doi: 10.1177/0962280206074466.
34. Trygged S, Ahacic K, Kåreholt I. Income and education as predictors of return to working life among younger stroke patients. BMC Public Health. 2011;11:742. doi: 10.1186/1471-2458-11-742.
CLINICAL PERSPECTIvESurvival after out-of-hospital cardiac arrest has improved in multiple sites during the past decade. Nevertheless, little is known about functional recovery in survivors and factors associated with this recovery during this period. In Denmark dur-ing 2001–2011, >75% (N=610) of working-age 30-day survivors were capable of returning to work. Among the survivors who returned to work, 75% (N=455) maintained employment without any sick leave relapse during the first 6 consecutive months after return to work and maintained annual income levels comparable to those seen before arrest. Importantly, along with improved survival, this proportion of survivors who returned to work increased from 66% (N=117) in 2001–2005 to 78% (N=338) in 2006–2011. Arrest in years 2006–2011, unlike in 2001–2005, was also significantly associated with return to work in our multivariable model. This association likely reflects multiple nationwide initiatives taken during our study period that we were not able to measure directly, including widespread resuscitation training, increase in emergency mobile units staffed with anesthesiologists or paramedics, dispatcher-assisted guidance of bystanders and guideline changes stressing the importance of chest compressions, and intensive care including therapeutic hypothermia. Among measurable variables, bystander- and emergency medical service–witnessed arrests were associated with return to work, and, when excluding emer-gency medical service–witnessed cases, so were arrests in which bystanders performed cardiopulmonary resuscitation. Our findings particularly speak to the potential of interventions that arrest witnesses, including bystanders, can initiate to improve outcomes. Conclusively, factors of known importance in survival were also associated with return to work.
Go to http://cme.ahajournals.org to take the CME quiz for this article.
by guest on May 11, 2015http://circ.ahajournals.org/Downloaded from
quartile 95 (20.8%) 99 (30.3%) 49.0% 0.49 0.33 0.72 0.001 missing 9 5 †Household income is indexed for the year 2010 and median in USD is 115,447 [IQR 79,679-149,629]. ‡Educational level: group 1
includes educational level of the International Standard Classification of Education 0-2, abbreviated ISCED 0-2: pre-primary education,
primary and lower secondary education; group 2 included ISCED 3 and 5 (ISCED 4 does not exist in Denmark): high school or vocational
education and short cycle tertiary education; and group 3 included ISCED 6-8: medium length tertiary education, bachelor, master and
5
Ph.D. degrees. §Time interval from recognition of OHCA (time of emergency call) to rhythm analysis by emergency medical service. *Abbreviations: OHCA, out-of-hospital cardiac arrest; CPR, cardiopulmonary resuscitation; EMS, emergency medical service; AED,
is online at: Circulation Information about subscribing to Subscriptions:
http://www.lww.com/reprints Information about reprints can be found online at: Reprints:
document. Permissions and Rights Question and Answer this process is available in the
click Request Permissions in the middle column of the Web page under Services. Further information aboutOffice. Once the online version of the published article for which permission is being requested is located,
can be obtained via RightsLink, a service of the Copyright Clearance Center, not the EditorialCirculationin Requests for permissions to reproduce figures, tables, or portions of articles originally publishedPermissions:
by guest on May 11, 2015http://circ.ahajournals.org/Downloaded from
