Embed Size (px)
Citation preview
Original Investigation
Employment of Patients Receiving Maintenance Dialysis andAfter Kidney Transplant: A Cross-sectional Study From Finland
Ilkka Helanterä, MD, PhD,1 Mikko Haapio, MD,1 Petri Koskinen, MD, PhD,1
Carola Grönhagen-Riska, MD, PhD,1 and Patrik Finne, MD, PhD1,2
Background: Associations between mode of renal replacement therapy and employment rate have notbeen well characterized.
Study Design: Cross-sectional registry analysis.Setting & Participants: The employment status of all prevalent 15- to 64-year-old dialysis and kidney
transplant patients in Finland at the end of 2007 (N � 2,637) was analyzed by combining data from the FinnishRegistry for Kidney Diseases with individual-level employment statistics of the Finnish government.
Predictor: Prevalence rate ratios (PRRs) of employment according to treatment modality with adjustment forage, sex, cause of end-stage renal disease (ESRD), duration of ESRD, and comorbid conditions wereestimated using Cox regression with a constant time at risk.
Outcome: Employment status of patients on dialysis therapy or after transplant.Measurements: Clinical data were collected from the Finnish Registry for Kidney Diseases, and employ-
ment data were acquired from Statistics Finland.Results: 19% of hemodialysis patients, 31% of peritoneal dialysis patients, and 40% of patients with a
functioning transplant were employed; the overall employment rate for the Finnish population aged 15-64 yearsis 67%. Home hemodialysis patients and those treated with automated peritoneal dialysis had employmentrates of 39% and 44%, respectively. In adjusted analysis, patients on home hemodialysis therapy (PRR, 1.87),on automated peritoneal dialysis therapy (PRR, 2.14), or with a kidney transplant (PRR, 2.30) had higherprobabilities of employment than in-center hemodialysis patients. Patients with type 1 or 2 diabetes as thecause of ESRD had the lowest probability of employment (PRR, 0.48-0.60 compared with glomerulonephritis).Patients aged 25-54 years more frequently were employed than those younger than 25 or older than 54 years.Sex did not predict employment. For transplant recipients, longer time since transplant was associated withhigher employment in addition to the mentioned factors.
Limitations: Cross-sectional design.Conclusions: Employment rate of home dialysis patients was similar to that of transplant recipients and
higher than that of in-center hemodialysis patients. Patients with diabetes were less likely to be employed.Am J Kidney Dis. 59(5):700-706. © 2012 by the National Kidney Foundation, Inc.
INDEX WORDS: Employment; hemodialysis; peritoneal dialysis; kidney transplantation.
Editorial, p. 598
The quality of life of patients with end-stage renaldisease (ESRD) is severely reduced, and pa-
tients on long-term dialysis therapy commonly experi-ence not only fatigue and deceased physical capabil-ity, but also impaired social and mental functioning.1-4
In addition, combining maintenance dialysis treat-
From the 1Deparment of Medicine, Division of Nephrology,Helsinki University Central Hospital; and 2Finnish Registry forKidney Diseases, Helsinki, Finland.
Received April 7, 2011. Accepted in revised form August 16,2011. Originally published online September 29, 2011.
Address correspondence to Ilkka Helanterä, MD, PhD, Depart-ment of Medicine, Division of Nephrology, Helsinki UniversityHospital, Haartmaninkatu 4, PO Box 372, FI-00029 HUS, Hel-sinki, Finland. E-mail: [email protected]
© 2012 by the National Kidney Foundation, Inc.0272-6386/$36.00
doi:10.1053/j.ajkd.2011.08.025700
ment with a regular job can be challenging. Therefore, itis not surprising that the frequency of employed long-term dialysis patients is low at 10%-30%, depending onthe population studied.4-7 Several factors are associatedwith increased employment during dialysis treatment,including higher education and higher earning power8,9
and employment before the start of dialysis therapy.7
Similarly, peritoneal dialysis (PD) as treatment modality,enabling ambulatory treatment with flexible schedules,has been associated with a higher employment rate insome,9-11 but not all, studies.8
A successful kidney transplant restores kidney func-tion and is associated with improved quality of lifecompared with maintenance dialysis therapy.12 Fre-quency of employment also is higher in transplant recipi-ents compared with long-term dialysis patients and isestimated to vary from 18%-82%, depending on trans-plant populations and definitions of employment sta-tus.13 Pretransplant employment is considered the mostimportant predictor of posttransplant employment.14-16
The aim of this cross-sectional study was to investi-
gate the employment rate across different modes ofAm J Kidney Dis. 2012;59(5):700-706
Employment During RRT
renal replacement therapy (RRT) in dialysis and kid-ney transplant patients in Finland and evaluate otherfactors associated with employment.
METHODSThe cross-sectional cohort of this study included all Finnish
patients aged 15-64 years either receiving long-term dialysistherapy or with a functioning kidney transplant on December 31,2007 (N � 2,637). Data were obtained from the Finnish Registryfor Kidney Diseases, which has an estimated 97%-99% coverageof all patients accepted for RRT (dialysis or kidney transplant) inFinland since 1965. The registry is maintained by the FinnishKidney and Liver Association and financed by the Finnish govern-ment. Written informed consent and permission to use the dataanonymously in registry reports and for research purposes wereobtained from all patients when entering the registry. The follow-ing data were obtained from the registry: patient age, sex, cause ofESRD (type 1 or 2 diabetes separately, glomerulonephritis, poly-cystic kidney disease, chronic pyelonephritis, amyloidosis, other,or unknown), current and previous modality of RRT (continuousambulatory PD [CAPD]; automated PD [APD]; in-center hemodi-alysis [HD], a group which also includes in-center hemodiafiltra-tion patients; home HD; or transplant) start date of RRT, time oflast change in treatment modality, and the presence of 9 comorbidconditions (history of hypertension, angina pectoris, myocardialinfarction, coronary bypass operation, arteriosclerosis, arterioscle-rotic surgery, limb amputation, left ventricular hypertrophy, orcardiac decompensation). Data for comorbid conditions were notavailable for patients who started RRT before 2000.
Information about the employment status of patients on Decem-ber 31, 2007, was acquired by database linkage with the annuallyupdated individual-level employment data of Statistics Finland.Statistics Finland is a public authority that has the right to collectdata for statistical purposes by virtue of the data supply obligationprescribed in Finnish law. The patient-level database linkage waspossible because of the Finnish system of a unique social securitycode for all citizens. Data for employment statistics of StatisticsFinland are derived from �40 administrative and statistical datafiles, from which the main type of activity of the population isconcluded. The most important of these data files are The Popula-tion Information System of the Population Register Centre, Regis-ters of the Tax Administration, Employment Registers of theFinnish Centre for Pensions, the State Treasury and the LocalGovernment Pensions Institution, Statistics Finland’s Register ofEnterprises and Establishments, Pensioner Registers of the SocialInsurance Institution and the Finnish Centre for Pensions, Ministryof Labour’s Register of Jobseekers, Statistics Finland’s Register ofCompleted Education and Degrees, Statistics Finland’s StudentRegister, and The Conscripts Register of the General Staff of theArmed Forces. We divided employment status into employed ornot employed. The latter group includes unemployed persons andpersons outside the labor force (homemakers, students, militaryconscripts, those on disability pension, retired, or unknown).
Differences in variable distributions between groups were ana-lyzed using nonparametric Kruskal-Wallis 1-way analysis (continu-ous variables) or �2 test (categorical variables). Statistical differ-ences detected by the Kruskal-Wallis test were analyzed furtherwith the Dunn test. Nonparametric statistics were applied becauseall distributions were not normal. The association between patientcharacteristics and employment was analyzed using a multivariateCox regression model with a constant time at risk. Employmentstatus was the binary outcome. It has been shown that when aconstant risk period is assigned to every person in the studypopulation, the hazard rate ratio estimated by Cox regression
equals the prevalence rate ratio (PRR) in cross-sectional studies.17Am J Kidney Dis. 2012;59(5):700-706
We used boot-strapping (with 5,000 bootstrap samples) to calcu-late robust confidence intervals (CIs) of the PRRs given by the Coxregression. Explanatory variables were treatment modality (in-center HD as the reference group), age (categorized as 15-24,25-34, 35-44, 45-54, and 55-64 years as the reference group), sex,cause of ESRD (glomerulonephritis as the reference group), sumof comorbid conditions from 0-8 (excluding hypertension becauseit was present in most patients), and time from start of RRT (years).When patients with a functioning transplant were analyzed sepa-rately, previous modality of treatment and time since transplant(years) also were included as variables in the model. Becausecomorbidity data were not available for all patients, the model wasperformed both with and without these variables. All first-degreeinteractions between treatment modality and the other explanatoryvariables were tested. Calculations were performed with PASWStatistics software (version 18.0.3; IBM Corp, www.spss.com).Two-sided P � 0.05 was considered statistically significant.
RESULTS
Patients included in the study are described inTables 1 and 2, organized by treatment modality(in-center HD [this group also includes hemodiafiltra-tion patients], home HD, APD, CAPD, or transplant).A total of 607 patients were categorized as HD pa-tients; this includes 550 individuals denoted as in-center HD patients (comprising 480 who were receiv-ing HD therapy and 70 treated with hemodiafiltration)and 57 patients performing home HD. Of 212 PDpatients, 74 patients were on CAPD therapy and 138patients were on APD therapy. Patients on in-centerHD therapy were significantly older than patients onother treatment modalities (P � 0.001). Comparedwith patients on HD or PD therapy, proportions ofpatients with glomerulonephritis and chronic pyelone-phritis as a cause of ESRD were larger for transplantrecipients, whereas type 1 diabetes was more commonin patients on PD therapy and type 2 diabetes wasmore common in patients on in-center HD therapy. Theproportion of patients with polycystic kidney diseasewas significantly smaller in PD compared with HD ortransplant patients. The number of comorbid conditionswas lower in transplant patients compared with HD orPD patients (P � 0.001). Detailed descriptions of comor-bid conditions in patients using different treatment mo-dalities are listed in Table 2. Patients on HD therapy hadbeen on RRT longer than patients on PD therapy (P �0.002), but shorter than transplant patients (P � 0.001).For transplant recipients, median time since transplantwas 7.0 (25th-75th percentile, 3.0-12.0) years.
A significantly higher proportion of transplant re-cipients (40%) were employed compared with pa-tients on HD (19%) or PD therapy (31%; P � 0.001),and a significantly higher proportion of patients on PDtherapy were employed compared with HD patients(P � 0.001). Employment rates were 39% for APDpatients, 16% for CAPD patients, and 44% for homeHD patients (P � 0.001). In comparison, the employ-
ment rate on December 31, 2007, in the entire Finnish701
Helanterä et al
population (aged 15-64 years) was 67.1%.18 Employ-ment rates according to age category in dialysis andtransplant patients and the entire Finnish populationare shown in Fig 1.
Table 3 lists results of the statistical model predict-ing the probability of employment in patients withavailable comorbidity data. Data for comorbid condi-tions were available for 1,408 individuals (192 of 212patients on PD therapy, 502 of 607 patients on HDtherapy, and 714 of 1,818 transplant recipients). Whenadjusting for age, sex, cause of ESRD, number ofcomorbid conditions, and time since start of RRT,patients on home HD (PRR, 1.87) and APD therapy(PRR, 2.14) and those with a functioning transplant(PRR, 2.30) had a higher probability of employmentcompared with the reference group of in-center HDpatients. The probability of employment of patientson CAPD therapy did not differ from the reference
Table 1. Patient Charact
In-center HD(n � 550)
HomeHD (n � 57)
Age (y) 52 � 10 47 � 10
Sex (M/F) 351/199 41/16
Cause of ESRDGlomerulonephritis 81 (14.7) 14 (24.6)Polycystic disease 57 (10.4) 11 (19.3)Diabetes type 1 124 (22.5) 9 (15.8)Diabetes type 2 97 (17.6) 4 (7)Chronic pyelonephritis 18 (3.3) 4 (7)Amyloidosis 21 (3.8) 0 (0)Other 85 (15.5) 9 (15.8)Unknown 67 (12.2) 6 (10.5)
No. of comorbid conditions 0.8 � 1.3 0.5 � 1.1
No. of patients employed 87 (15.8) 25 (44)
Years since start of RRT 4.5 � 5.9 5.0 � 6.3
Note: N � 2637. Continuous variables shown as mean � standAbbreviations: APD, automated peritoneal dialysis; CAPD,
disease; HD, hemodialysis; RRT, renal replacement therapy.
Table 2. Comorbid Conditions
In-center HD(n � 455)
HomeHD (n � 47)
Angina pectoris 43 (9.5) 0 (0)
Myocardial infarction 41 (9.0) 4 (8.5)
Coronary bypass operation 38 (8.4) 6 (12.8)
Arteriosclerosis 51 (11.2) 1 (2.1)
Arteriosclerotic surgery 27 (5.9) 2 (4.3)
Limb amputation 25 (5.5) 1 (2.1)
Left ventricular hypertrophy 124 (27.3) 8 (17.0)
Cardiac decompensation 30 (6.6) 1 (2.1)
Note: Values shown as number (percentage).
Abbreviations: APD, automated peritoneal dialysis; CAPD, continuo702
group. Type 1 or 2 diabetes (PRRs, 0.60 and 0.48,respectively) as cause of ESRD predicted a clearlylower probability of employment compared with thereference group of patients with glomerulonephritis.Age categories between 25-54 years predicted a higheremployment rate compared with the oldest age groupof 55- to 64-year-olds. Increasing number of comor-bid conditions and longer duration of RRT predicted alower probability of employment. No statisticallysignificant interactions were found between treatmentmodality and any other variable in the model.
When the model was calculated in the entire patientpopulation (N � 2,637) without data for comorbidconditions (missing from a subset of patients), allsignificant predictors in the previous model persisted.PRRs for other factors remained in the similar range:home HD (PRR, 1.84; 95% CI, 1.27-2.52; P �0.001), APD (PRR, 2.23; 95% CI, 1.71-2.93; P �
cs by Treatment Modality
n � 138) CAPD (n � 74)Functioning
Transplant (n � 1,818) P
� 11 49 � 9 49 � 11 �0.001
/47 43/31 1,106/712 0.2
�0.001(21.7) 9 (12.2) 510 (28.1)(4.3) 4 (5.4) 251 (13.8)(39.1) 29 (39.2) 436 (24)(10.1) 10 (13.5) 31 (1.7)(5.8) 1 (1.4) 151 (8.3)(0) 1 (1.4) 29 (1.6)(13) 15 (20.3) 342 (18.8)
(5.8) 5 (6.8) 68 (3.7)
� 1.6 1.0 � 1.2 0.3 � 0.7 �0.001
(39.1) 12 (16.2) 724 (40) �0.001
� 5.1 2.1 � 2.5 11.5 � 7.8 �0.001
eviation; categorical variables, as number (percentage).nuous ambulatory peritoneal dialysis; ESRD, end-stage renal
atients by Treatment Modality
n � 122) CAPD (n � 70)Functioning
Transplant (n � 714) P
(9.8) 10 (14.3) 21 (2.9) �0.001
(10.7) 8 (11.4) 22 (3.1) �0.001
(9.8) 7 (10.0) 29 (4.1) 0.003
(9.0) 11 (15.7) 16 (2.2) �0.001
(4.9) 4 (5.7) 10 (1.4) �0.001
(3.3) 9 (12.9) 6 (0.8) �0.001
(26.2) 20 (28.6) 122 (17.1) �0.001
(9.8) 0 (0) 3 (0.4) �0.001
eristi
APD (
48
91
306
541480
188
0.8
54
3.5
ard dconti
of P
APD (
12
13
12
11
6
4
32
12
us ambulatory peritoneal dialysis; HD, hemodialysis.
Am J Kidney Dis. 2012;59(5):700-706
Employment During RRT
0.001), or transplant as treatment modality (PRR,2.11; 95% CI, 1.75-2.59; P � 0.001); type 1 or type 2diabetes as cause of ESRD (PRRs, 0.47 [95% CI,0.40-0.54; P � 0.001] and 0.42 [95% CI, 0.24-0.64;P � 0.001], respectively); and age of 25-34 (PRR,2.61; 95% CI, 2.16-3.13; P � 0.001), 35-44 (PRR,2.87; 95% CI, 2.48-3.34; P � 0.001), or 45-54 years(PRR, 3.23; 95% CI, 1.93-2.56; P � 0.001).
The multivariate model also was constructed sepa-rately including only patients on dialysis therapy (exclud-ing transplant recipients) with similar results for treat-ment modality (home HD and APD) and age (25-54compared with 55-64 years) predicting employment,and type 1 or 2 diabetes as cause of ESRD predictinglower probability of employment (data not shown).
Employment of kidney transplant recipients withavailable comorbidity data was analyzed separately ina multivariate model, listed in Table 4. Of 714 trans-plant patients with available comorbidity data,90 (12.6%) were on CAPD therapy before transplant,192 (26.9%) were on APD therapy, 367 (51.4%) wereon in-center HD or hemodiafiltration therapy, and62 (8.7%) were on home HD therapy before trans-plant. In this multivariate model, home HD (PRR,2.14) or APD (PRR, 1.69) as dialysis modality beforetransplant predicted employment (Table 4). Type 1 or2 diabetes (PRRs, 0.57 and 0.42, respectively) ascause of ESRD predicted lower probability of employ-ment. Longer time since transplant (OR, 1.05 per 1
Figure 1. Employment rate (%) at differ-ent age categories in the entire Finnish popu-lation,17 patients with a functioning trans-plant, patients on peritoneal dialysis therapy,and patients on hemodialysis therapy.
year) and age of 25-54 years also predicted employ-
Am J Kidney Dis. 2012;59(5):700-706
ment. Number of comorbid conditions was associatedwith a lower probability of employment after transplant(PRR, 0.80 per 1 comorbid condition increase). Nosignificant interactions were recorded between treatmentmodality and any other variable in the model.
When the model was calculated in a patient popula-tion (n � 1,818) including all transplant recipients with-out data for comorbid conditions (which were missingfor a subset of patients), all significant predictors per-sisted. PRRs for other factors remained in a similarrange: home HD (PRR, 1.88; 95% CI, 1.52-2.29; P �0.001) or APD (PRR, 1.63; 95% CI, 1.41-1.88; P �0.001) as pretransplant treatment modality, type 1 ortype 2 diabetes as cause of ESRD (PRRs, 0.46 [95% CI,0.38-0.54; P � 0.001] and 0.41 [95% CI, 0.11-0.79; P �0.02], respectively), and age of 25-34 (PRR, 2.42; 95%CI, 1.96-2.99; P � 0.001), 35-44 (PRR, 2.96; 95% CI,2.53-3.52; P � 0.001), or 45-54 years (PRR, 2.23; 95%CI, 1.90-2.62; P � 0.001).
DISCUSSION
In this cross-sectional registry analysis combiningdata from the Finnish Registry for Kidney Diseaseswith the official individual-level employment statis-tics of Finland, we show that a functioning transplant,APD therapy, and home HD therapy were connectedwith a higher employment rate in patients with ESRDcompared with in-center HD therapy. Furthermore,APD or home HD as the last treatment modality
before transplant was associated with more frequent703
Helanterä et al
employment after transplant. Diabetes as a cause ofESRD was associated with a lower probability ofemployment.
Similar to our results, earlier cross-sectional studieshave shown an association between PD therapy andhigher rates of employment.9-11,19 However, patientswho are employed when maintenance dialysis therapy isinitiated probably tend to choose PD, especially APD, asthe treatment modality because it enables ambulatorytreatment with flexible schedules (eg, at home at nightwith APD). This may explain why the employment rateof patients on CAPD therapy was not higher comparedwith patients on in-center HD therapy. Accordingly,evidence suggests that treatment modality itself does nothave an impact on employment rate, but employmentstatus influences the choice of treatment modality.7,9 In aprospective longitudinal study by van Manen et al7
analyzing changes in employment status during the firstyear of dialysis therapy, treatment modality (HD vs PD)did not predict loss of work and most patients who wereemployed before starting dialysis therapy were able to
Table 3. Multivariate Model of Factors Predicting Employmentof RRT Patients
Variable PRR (95% CI) P
Treatment modalitya �0.001Home HD (n � 47) 1.87 (1.26-2.64) 0.001CAPD (n � 70) 0.73 (0.34-1.26) 0.3APD (n � 122) 2.14 (1.59-2.83) �0.001Functioning transplant
(n � 714)2.30 (1.85-2.92) �0.001
Age groupb �0.00115-24 y 1.10 (0.61-1.69) 0.725-34 y 2.32 (1.80-2.94) �0.00135-44 y 2.36 (1.90-2.92) �0.00145-54 y 1.84 (1.50-2.24) �0.001
Male sex 0.91 (0.79-1.08) 0.3
Cause of ESRDc �0.001Polycystic disease 1.15 (0.92-1.43) 0.2Diabetes type 1 0.60 (0.49-0.74) �0.001Diabetes type 2 0.48 (0.26-0.74) 0.003Chronic pyelonephritis 0.85 (0.56-1.17) 0.4Amyloidosis 0.66 (0.24-1.13) 0.2Other 0.87 (0.68-1.10) 0.3Unknown 0.91 (0.69-1.15) 0.4
No. of comorbid conditions 0.84 (0.74-0.93) 0.002
Time since start of RRT (y) 0.91 (0.88-0.95) �0.001
Note: N � 1,408.Abbreviations: APD, automated peritoneal dialysis; CAPD,
continuous ambulatory peritoneal dialysis; CI, confidence inter-val; ESRD, end-stage renal disease; HD, hemodialysis; PRR,prevalence rate ratio; RRT, renal replacement therapy.
aReference is in-center HD therapy (n � 455).bReference is 55-64 years.cReference is glomerulonephritis.
maintain their work. Employment rate at the start of
704
dialysis therapy was already low, and very few unem-ployed patients became employed during the first year ofdialysis therapy.7 In our cross-sectional study, data foremployment status before starting dialysis therapy werenot available.
In a manner similar to PD, home HD also offersfreedom from treatment schedules and possibility forout-of-work-hours dialysis, making it a suitable optionfor employed patients. Only limited data are available,but employment rate has been suggested to be higher inhome HD patients.20 In addition, the possibility of homeHD in a dialysis facility has been associated with higheremployment rates in a US national survey of dialysisunits.6 In our analysis, we could confirm that home HDpredicted a higher employment rate, and furthermore,transplant recipients who had been on home HD therapybefore transplant were more likely to be employed.
In our study, employment rates were similar amongAPD, home HD, and transplant patients and were 39%-44%. Comparison of employment rates between earlierstudies is difficult because there are differences in theanalyzed dialysis patient populations (prevalent dialysispatients vs patients starting dialysis therapy), age of
Table 4. Multivariate Model of Factors Predicting EmploymentAfter Kidney Transplant
Variable PRR (95% CI) P
Treatment modality beforetransplanta
�0.001
Home HD (n � 62) 2.14 (1.68-2.74) �0.001CAPD (n � 90) 1.33 (0.96-1.77) 0.06APD (n � 192) 1.69 (1.37-2.07) �0.001
Age groupb �0.00115-24 y 1.13 (0.61-1.78) 0.625-34 y 1.99 (1.42-2.75) �0.00135-44 y 2.08 (1.63-2.73) �0.00145-54 y 1.85 (1.45-2.40) �0.001
Male sex 0.94 (0.77-1.14) 0.5
Cause of ESRDc 0.03Polycystic disease 1.07 (0.82-1.39) 0.6Diabetes type 1 0.57 (0.44-0.73) �0.001Diabetes type 2 0.42 (0.09-0.84) 0.03Chronic pyelonephritis 0.80 (0.47-1.17) 0.3Amyloidosis 0.63 (0.19-1.07) 0.1Other 0.80 (0.59-1.06) 0.1Unknown 0.94 (0.68-1.26) 0.7
No. of comorbid conditions 0.80 (0.65-0.94) 0.02
Time since transplant (y) 1.05 (1.01-1.10) 0.03
Note: N � 714.Abbreviations: APD, automated peritoneal dialysis; CAPD,
continuous ambulatory peritoneal dialysis; CI, confidence inter-val; ESRD, end-stage renal disease; HD, hemodialysis; PRR,prevalence rate ratio.
aReference is in-center HD therapy (n � 370).bReference is 55-64 years.
cReference is glomerulonephritis.Am J Kidney Dis. 2012;59(5):700-706
Employment During RRT
patients, and importantly, the social security systemsbetween different countries. In other studies, the employ-ment rate in PD patients has been 28%-50%,7-9,19 whichis similar to our results. Similarly, other studies havereported lower rates of employment in HD patientscompared with PD patients at 11%-32%,4,8,9,19 whereasthe employment rate in kidney transplant recipients hasbeen 30%-70%.13 Interestingly, employment rates didnot differ betweenAPD, home HD, or transplant patientsin our study, further supporting the idea that employmentstatus determines the choice of RRT and not vice versa.These findings also suggest that home dialysis modali-ties may be associated with the same degree of rehabili-tation as kidney transplant.
Employment rate probably also is affected by theFinnish social security system, in which governmentsubsidies are provided to the unemployed or chronicallyill, and in some cases, low-paid employment may notincrease yearly income substantially compared with so-cial benefits. Thus, income level and occupation alsoprobably are important predictors of employment in thechronically ill. In comparison to other patient popula-tions studied in Finland, the employment rate 1 year aftercoronary artery bypass grafting in patients younger than60 years and not retired preoperatively was 47%,21
whereas in a registry analysis of cancer survivors, theemployment rate was 64% at 2-3 years after cancerdiagnosis.22 In these studies, higher education, previousemployment, and younger age positively predicted em-ployment, whereas manual labor was associated nega-tively with employment.21,22 Data for occupation oreducation level of patients were not included in theavailable registry data in the present study, limiting ouranalyses.
Despite the strong association, it is not justified basedon this cross-sectional analysis to state that there wouldbe a causal effect of treatment modality on employmentrate. For many patients in poor condition and disabledpatients, in-center HD is the only suitable dialysis modal-ity. Moreover, employed patients who are better off maybe more likely to select a home treatment modality.Adjusting for confounding factors, we explored the roleof several cardiovascular comorbid conditions andshowed that the association between treatment modalityand employment status was independent of the numberof comorbid conditions. However, our data did notinclude some comorbid condition that may affect abilityto work, such as infections, inflammatory diseases, andmalignancies. Thus, we could not exclude them as con-founding factors. However, our model was adjusted forcause of ESRD, which was associated with employmentstatus because patients with diabetes were less likely tobe employed. Finland has a high incidence of type 1diabetes, which was the cause of ESRD in 25% of
patients in the present study, affecting the employmentAm J Kidney Dis. 2012;59(5):700-706
rate compared with other populations. One importantfactor not included in the analysis, possibly explainingthe lower employment rate in patients with diabetes,could be visual impairment, which commonly is presentin diabetic patients with ESRD.
Several factors associated with the employment ofkidney transplant recipients have been identified in pre-vious studies, most importantly employment status be-fore transplant.14-16 In addition, longer time from trans-plant,15,23 not having diabetes,14 lack of cardiovascularcomorbid conditions,23 and PD therapy before trans-plant24 have been connected with a higher employmentrate. In our registry analysis including a large number oftransplant recipients, we could confirm the findings ofthe association of pretransplant PD therapy (APD) andlonger time from transplant with a higher probability ofemployment after transplant. The positive associationbetween time from transplant and employment rate pos-sibly could be explained by the fact that many years aftertransplant, patients are more likely to have recoveredfrom the operation and early complications and areadapted to the state of immunosuppression. In patientswho have not returned to dialysis therapy or died, thetransplant probably is functioning well. Similarly, pa-tients with diabetes were less likely to be employed aftertransplant, and number of comorbid conditions alsopredicted a lower rate of employment. In addition, as anovel finding, we could show that home HD therapybefore transplant is associated with a higher rate ofemployment after transplant. However, although datafor pretransplant employment were not available, itis most likely that our findings regarding pretrans-plant home dialysis modalities reflect employmentstatus before transplant.
In our study, age also was a significant predictor ofemployment in both dialysis patients and transplantrecipients. Patients in the youngest (15-24 years) andoldest (55-64 years) age categories were less likely to beemployed. Data acquired from Statistics Finland in-cluded information for only employment status, and not,for example, student status, which could explain the lowrate of employment in the youngest age group. Similarly,in the oldest age group, retirement due to age explainslower employment to some extent; at present, personsstarting their work career in Finland are entitled to retirebetween 63 and 68 years, and in older-age cohorts, theage limit has been even lower. In 2007, the median ageof retirement due to age in Finland was 60.1 years.18 Wedid not find an association between employment andsex, whereas in the entire Finnish population in 2007, theemployment rate was slightly higher in male comparedwith female persons.18
Our study has some limitations resulting from thedesign of the study. First, the only medical data available
for patients were those recorded in the Finnish Registry705
Helanterä et al
of Kidney Diseases. Therefore, we could not analyze theimpact of several factors suggested as important predic-tors of employment in patients with ESRD in otherstudies. These factors include employment status beforethe initiation of dialysis therapy or before transplant7,16
and education.8,9 In addition, because this was a cross-sectional study, no follow-up data for employment wereavailable and no conclusions about causality of thefactors associated with employment can be drawn. How-ever, the strength of the study is the almost completecoverage of patients on RRTs in the entire country ofFinland. Moreover, because employment data were ac-quired from official government statistics, these data canbe regarded as highly reliable and comprehensive. Incomparison, most earlier studies used employment datathat were self-reported and with variable percentages ofreply, thus possibly biasing results.4,7,8,15,24-26
In conclusion, in this cross-sectional registry analysis,we showed that home dialysis modalities are associatedwith a higher employment rate not only during mainte-nance dialysis therapy, but also after transplant, and thatthe employment rate of patients on home dialysis modali-ties is similar to that for transplant patients. In addition,patients with diabetes or cardiovascular comorbid condi-tions are less likely to be employed.
ACKNOWLEDGEMENTSSupport: This study was funded by the Academy of Finland (to
Dr Helanterä) and Finnish Kidney and Liver Association (FinnishRegistry for Kidney Diseases).
Financial Disclosure: The authors declare that they have noother relevant financial interests.
REFERENCES1. Merkus MP, Jager KJ, Dekker FW, de Haan RJ, Boeschoten
EW, Krediet RT. Physical symptoms and quality of life in patientson chronic dialysis: results of the Netherlands Cooperative Studyon Adequacy of Dialysis (NECOSAD). Nephrol Dial Transplant.1999;14(5):1163-1170.
2. Merkus MP, Jager KJ, Dekker FW, Boeschoten EW, StevensP, Krediet RT. Quality of life in patients on chronic dialysis:self-assessment 3 months after the start of treatment. TheNECOSAD Study Group. Am J Kidney Dis. 1997;29(4):584-592.
3. Moore GE, Parsons DB, Stray-Gundersen J, et al. Uremicmyopathy limits aerobic capacity in hemodialysis patients. Am JKidney Dis. 1993;22(2):277-287.
4. Molsted S, Aadahl M, Schou L, Eidemak I. Self-rated healthand employment status in chronic haemodialysis patients. ScandJ Urol Nephrol. 2004;38(2):174-178.
5. Kutner NG, Brogan D, Fielding B. Employment status andability to work among working-age chronic dialysis patients. Am JNephrol. 1991;11(4):334-340.
6. Kutner N, Bowles T, Zhang R, Huang Y, Pastan S. Dialysisfacility characteristics and variation in employment rates: a na-tional study. Clin J Am Soc Nephrol. 2008;3(1):111-116.
7. van Manen JG, Korevaar JC, Dekker FW, Reuselaars MC,Boeschoten EW, Krediet RT; NECOSAD Study Group. Nether-
lands Cooperative Study on Adequacy of Dialysis: changes in706
employment status in end-stage renal disease patients during theirfirst year of dialysis. Perit Dial Int. 2001;21(6):595-601.
8. Curtin RB, Oberley ET, Sacksteder P, Friedman A. Differ-ences between employed and nonemployed dialysis patients. Am JKidney Dis. 1996;27(4):533-540.
9. Hirth RA, Chernew ME, Turenne MN, Pauly MV, Orzol SM,Held PJ. Chronic illness, treatment choice and workforce participa-tion. Int J Health Care Finance Econ. 2003;3(3):167-181.
10. Wolcott DL, Nissenson AR. Quality of life in chronicdialysis patients: a critical comparison of continuous ambulatoryperitoneal dialysis (CAPD) and hemodialysis. Am J Kidney Dis.1988;11(5):402-412.
11. Julius M, Kneisley JD, Carpentier-Alting P, HawthorneVM, Wolfe RA, Port FK. A comparison of employment rates ofpatients treated with continuous ambulatory peritoneal dialysis vsin-center hemodialysis (Michigan End-Stage Renal Disease Study).Arch Intern Med. 1989;149(4):839-842.
12. Cameron JI, Whiteside C, Katz J, Devins GM. Differencesin quality of life across renal replacement therapies: a meta-analytic comparison. Am J Kidney Dis. 2000;35(4):629-637.
13. van der Mei SF, Krol B, van Son WJ, de Jong PE, GroothoffJW, van den Heuvel WJ. Social participation and employmentstatus after kidney transplantation: a systematic review. Qual LifeRes. 2006;15(6):979-994.
14. Matas AJ, Lawson W, McHugh L, et al. Employmentpatterns after successful kidney transplantation. Transplantation.1996;61(5):729-733.
15. Markell MS, DiBenedetto A, Maursky V, et al. Unemploy-ment in inner-city renal transplant recipients: predictive and so-ciodemographic factors. Am J Kidney Dis. 1997;29(6):881-887.
16. Raiz L. The transplant trap: the impact of health policy onemployment status following renal transplantation. J Health SocPolicy. 1997;8(4):67-87.
17. Barros AJ, Hirakata VN. Alternatives for logistic regressionin cross-sectional studies: an empirical comparison of models thatdirectly estimate the prevalence ratio. BMC Med Res Methodol.2003;3:21-33.
18. Statistics Finland. Official statistics of the Finnish govern-ment 2007. www.stat.fi. Accessed December 12, 2010.
19. Rubin HR, Fink NE, Plantinga LC, Sadler JH, Kliger AS,Powe NR. Patient ratings of dialysis care with peritoneal dialysisvs hemodialysis. JAMA. 2004;291(6):697-703.
20. Agar JW. Nocturnal haemodialysis in Australia and NewZealand. Nephrology (Carlton). 2005;10(3):222-230.
21. Hällberg V, Palomäki A, Kataja M, Tarkka M. Return towork after coronary artery bypass surgery. A 10-year follow-upstudy. Scand Cardiovasc J. 2009;43(5):277-284.
22. Taskila-Åbrandt T, Martikainen R, Virtanen SV, Pukkala E,Hietanen P, Lindbohm ML. The impact of education and occupa-tion on the employment status of cancer survivors. Eur J Cancer.2004;40(16):2488-2493.
23. van der Mei SF, Groothoff JW, van Sonderen EL, van denHeuvel WJ, de Jong PE, van Son WJ. Clinical factors influencingparticipation in society after successful kidney transplantation.Transplantation. 2006;82(1):80-85.
24. van der Mei SF, van Son WJ, van Sonderen EL, de Jong PE,Groothoff JW, van den Heuvel WJ. Factors determining socialparticipation in the first year after kidney transplantation: a prospec-tive study. Transplantation. 2007;84(6):729-737.
25. Bohlke M, Marini SS, Gomes RH, et al. Predictors ofemployment after successful kidney transplantation—a population-based study. Clin Transplant 2008;22(4):405-410.
26. De Baere C, Delva D, Kloeck A, et al. Return to work andsocial participation: does type of organ transplantation matter?
Transplantation. 2010;89(8):1009-1015.Am J Kidney Dis. 2012;59(5):700-706
