Differences between employed and nonemployed dialysis patients

Differences Between Employed and Nonemployed Dialysis Patients Roberta Braun Cut-tin, PhD, Edith T. Oberley, MA, Paulette Sacksteder, and Aaron Friedman, MD

BS,

0 Three hundred fifty-nine chronic dialysis patients (SE employed and 274 nonemployed) were surveyed to identify/ verify those characteristics which differentiate between employed versus nonemployed status. Education emerged as a significant correlate of employment, as noted by previous investigators, whereas, unlike previous research, neither mode of dialysis, length of time on dialysis, number of comorbid conditions, nor cause of renal failure (eg, diabetes) were associated with employment status. Measures of functional status (MOS SF-20 and Kamofsky) were positively associated with employment. Furthermore, patients’ perceptions that their health limited the type and amount of work that they could do were negatively associated with employment. In addition, using a series of de novo items, we found subjects’ beliefs about dialysis patients’ ability to work to be a “self-fulfilling prophecy” with regard to employment status. That is, patients who themselves believed that dialysis patients should work and had this notion reinforced by significant others were more likely to be employed. Interestingly, 21% of unemployed patients reported that they were both able to work and wanted to return to work. Because it is consistently reported that only a small percentage of dialysis patients are employed, targeting the patients who are both willing and able to work for vocational rehabilitation might significantly increase the numbers of employed dialysis patients. 0 1996 by the National Kidney Foundation, Inc. INDEX WORDS: work; vocational

Rehabilitation; status.

dialysis

patients;

employment;

end-stage

S

Journal

of Kidney

Diseases,

Vol 27, No 4 (April),

disease;

attitudes;

functional

status;

sample of dialysis patients focusing on factors that have been previously identified as likely to affect their employment.

INCE the enactment of Medicare coverage for end-stage renal disease (ESRD) patients in 1973, the possibilities for the rehabilitation and employment of dialysis patients have appeared to decline as this population has aged and now experiences more concurrent disease.lm4 On the other hand, medical, technologic, and pharmaceutical advances have improved well-being for a great many patients. As a result, many renal patients who were previously handicapped by fatigue are able to enjoy levels of energy and vigor similar to their pre-ESRD levels.5-7 Because adequate energy is a necessary condition for employment, increases in the numbers of dialysis patients who are employed is also a reasonable expectation. However, clinical reports continue to demonstrate that only a small percentage of working-age dialysis patients are currently employed.*-” A whole range of obstacles to employment, including health-related barriers, economic barriers, and attitudinal barriers, have an impact on renal patients’ employment.2,‘.*.’ ’ Although many of these impediments seem as if they are universal for all dialysis patients, there continues to be that small percentage of dialysis patients who are employed. Thus, identifying distinguishing characteristics between dialysis patients who are employed and those who are not seems a worthwhile goal. ‘* In this study, we surveyed a diverse national American

renal

MATERIALS

AND

METHODS

Sample Design Dialysis centers served as the sampling units in this study’s design. Facilities were randomly selected from each of the 10 geographic areas delineated in the National Listing of Medicare Providers Furnishing Kidney Dialysis and Transplant Services. To insure as diverse a subject population as possible, centers in each area were stratified by rural (population -=lO,OOO) versus urban location. The final sample of dialysis facilities consisted of 31 centers, 18 urban and 13 rural. Survey data were collected from 359 patients currently undergoing dialysis treatments. Patients were asked to participate in the study if they were between the ages of 18 and 62 years and were able to independently complete the survey.

From the Medical Education Institute, University Research Park, Madison, WI; and the Department of Pediatric Nephrology, University of Wisconsin, Madison, WI. Received September 26, 1995; accepted in revised form November 28, 1995. Supported by a clinical research grant from Amgen, Inc, Thousand Oaks, CA, Epogen Clinical Grants Program. The project was conducted through the Medical Education Institute of Madison, WI, a not-for-projit, private research organization. Data were collected from 359 dialysis patients in 31 dialysis centers located throughout the United States. Address reprint requests to Roberta Braun Curtin, PhD, 3520 Timber Lane, Cross Plains, WI 53528. 0 1996 by the National Kidney Foundation, Inc. 0272-6386/96/2704-OOIO$3.00/0 1996:

pp 533-540

533

CURTIN

534 Sixty-nine proached.

eligible

patients

declined

to participate

when

ap-

Measures Data were collected using a self-administered written questionnaire and one staff-completed measure. The patient selfadministered questionnaire included several items from a survey originally used in the US Renal Data System Special Study 4: EPO and Quality of Life,” which addressed demographic and some attitudinal information, a series of de novo items dealing with patients’ attitudes and the Medical Outcomes Study Short Form-20 (MOS SF-20). In addition to this questionnaire, the Kamofsky activity scale was completed for each respondent by the dialysis center’s social worker or dialysis nurse.

Demographic

and Attitudinal

Variables

Data were obtained regarding subjects’ age, gender, race, work history, current employment status, occupational prestige, cause of renal failure, comorbid conditions, length of time on dialysis, attitudes toward employment, perceptions of barriers to employment, and perceptions of health care professionals’ and families’ attitudes toward ESRD patients’ potential to work. The survey instrument was pretested on a pilot (n = 10) sample of dialysis patients with special attention directed at items addressing patients’ attitudes regarding barriers to employment and others’ expectations of their employability. Patients participating in the pretest phase initially completed the questionnaires independently. Following completion, they were interviewed by an experienced renal social worker to verify the face validity and content validity of the items. No problems were encountered, ie, patients understood the items and provided essentially the same answers in lengthy narrative form as had been provided in the written questionnaire itemsI Construct validity of the de novo items was also confirmed since findings regarding perceptions of barriers and others’ expectations were entirely consistent with previous research.14 Internal consistency of these items was demonstrated on the total sample with a Cronbach’s alpha of 0.76.”

Patient Functioning The Medical assess patients’

Outcomes Study level of functional

(SF-20)* was included to status, subjective feelings

* MOS total scores were calculated as follows: reverse coding was done so that in all cases higher scores indicate higher (better) functioning, all item scores were transformed linearly to a scale from 0 to 100, item mean substitution for any missing data was performed, and subjects’ individual item scores were summed to generate the final MOS total score. It is important to note that when separate chi-square analyses were performed for each of the six MOS SF-20 subscales, higher functioning was significantly positively associated with employment in every case: physical functioning, P = 0.000; role functioning, P = 0.000; social functioning, P = 0.000; mental health, P = 0.003; health perception, P = 0.000; and pain state, P = 0.001.

ET AL

of wellness/illness, current pain state, social functioning, and basic affective status.” It should be noted that psychometric assessments of the MOS SF-20, performed by its authors, yielded reliabilities well within acceptable range both for the total instrument and the six subscales. Validity analyses (item discrimination among scales, correlations between these measures and those in long form versions, and consistent patterns between present and previous research) speak to an acceptable level of instrument validity.16 In addition to the self-administered questionnaire, renal social workers or nurses in the selected facilities completed the Kamofsky Activity Scale to rate each patient’s functional status.” This measure was added to allow us to obtain an “outside” judgment and corroboration of the functional status evaluation (self-assessment) provided by the patients themselves.‘8 In this manner, both the validity of the previous measures and the accuracy and/or reliability of patient responses were verified. Although the strengths and weaknesses of the Kamofsky scale have been debated, its overall usefulness has been well established.‘.‘.6.7.‘8

Statistical Analysis Comparisons between employed and nonemployed subjects were performed for all measured categorical independent variables using chi-square analyses. Although continuous independent variables are also reported in Tables 1 to 4 in “categorical form,” they were analyzed as continuous variables using Student’s r-tests and are reported in this manner merely for ease of data presentation. Since employment status is a single-item, dichotomous variable, logistic regression was performed so that the multiple independent variables could be considered simultaneously. Categorical variables were re-coded as dummy varables for the purpose of these analyses. Coefficients, standard errors, odds ratios, and probabilities are presented for each variable in the model. The chi-square and probability for the overall model are also reported.

RESULTS

The overall patient population (n = 359) was almost evenly divided by gender: 49% women and 50% men. Subjects’ mean age was 43 t 10.6 years, a slightly younger group than in other recent studies of dialysis patients.2~8~‘9Of the 359 patients, 3 11 (87%) were receiving incenter hemodialysis, 42 were receiving peritoneal dialysis, and only three patients were receiving home hemodialysis (missing data = 3). As shown in Table 1, there was no significant difference found between working and nonworking subjects with regard to mode of dialysis or with regard to urban versus rural dialysis center location. Almost 24% of the subjects had less than a high school education, 40% were high school graduates, 24% had some college education, and

EMPLOYED

v NONEMPLOYED

DIALYSIS Table

PATIENTS

535

1. Demographic

Characteristics

(n = 359)

Nonemployed (n = 274)

Gender Male Female Age W 18-30 31-40 41-50 51-62 Race White Black Native American Hispanic Asian Treatment mode In-center hemodialysis Home hemodialysis Peritoneal dialysis Setting Urban Rural Education Some high school High school graduate Some college College graduate Postgraduate *Total n may vary due to missing T For the purposes of this study, or outside of the home.

Employedt (n = 85)

n*

%

rl*

%

126 145

46.0 53.0

50 35

58.8 41.2

36 73 81 72

13.1 26.6 29.6 26.3

13 19 27 25

15.3 22.4 31.8 29.4

127 101 22 14 6

46.3 37.0 8.1 5.2 2.2

49 27 2 1 5

57.6 31.8 2.4 1.2 5.9

238 3 30

86.9 1.1 11.1

73 0 12

85.9 0 14.1

208 66

75.9 24.0

73 12

85.8 14.1

81 111 61 10 8

29.6 40.4 22.6 3.7 3.0

4 31 27 14 8

4.7 36.5 32.1 16.7 9.5

Probability

Value

0.05

0.59

0.05

0.21

0.05

<0.0001

data. “employed”

was

defined

11% had graduated college. As shown in Table 1 and consistent with previous studies, a highly significant positive relationship was found between employment status and level of education. 8*20,2’Only 24% of the subjects were currently employed, either full or part time, whereas before dialysis 73% were employed.2,3*20 The measures used to estimate patients’ levels of health and functioning were length of time on dialysis, cause of renal failure, number of comorbid conditions, MOS SF-20 score, and Karnofsky score. As shown in Table 2, subjects had been on dialysis for a mean of 3.8 + 4.21 years. The two most common causes of renal disease were hypertension (26%) and diabetes (25%) with various other causes for renal failure accounting for the remainder of the cases (Table 2). Patients’ comorbid conditions ranged from 0 to 11 (mean,

as working

for pay,

either

full or part

time

and either

inside

2.65 + 1.99) with nearly 10% of the respondents suffering from six or more comorbid conditions. The possible range of the MOS total score was 0 to 2,000. As Table 2 shows, the MOS scores ranged from 20 to 1,960 with a mean score of 958 ? 390, while the Karnofsky ratings ranged from 40 to 100 with a mean score of 82 + 1.4).2,20 When nonemployed versus employed patients were compared on these health and functioning measures, only the MOS and Kamofsky scores proved to be significantly related to employment status, with higher scores in both cases associated with greater likelihood of employment. Patients’ assessment and understanding of health-related impediments to employment were also highly significantly associated with employment (Table 3). Employed patients reported that they were not limited by their health in the hours

536 Table

2. Health

and

Functioning

of Subjects

* Total

n may

vary

due to missing

ET AL

Probability

Values

(n = 359)

Nonemployed (n = 274)

Tlme on dialysis (mean, 3.83 i 4.21 yr) lO yr Cause of failure Diabetes Hypertension Lupus Polycystic Nephritis Congenital Scarlet fever/strep Drug toxicity Other/unknown Comorbidities (mean, 2.65 ? 1.99) O-l 2-5 26 MOS-20 score (mean, 958.4 t 390.5) 5630 631-850 851-l ,049 1,050-l ,265 >1,266 Karnofsky score (mean, 82 t 1.4) 40-60 70-80 90-l 00

CURTIN

Employed (n = 85)

n*

%

n*

%

58 136 46 25

22.1 51.1 17.1 9.5

25 39 12 8

30.1 46.5 13.2 9.6

72 72 20 14 16 13 10 8 43

26.7 26.3 7.0 5.2 5.6 4.8 3.7 3.0 15.6

18 19 4 8 5 4 7 0 19

21.2 22.4 4.7 9.4 5.9 4.7 8.2 0 22.3

89 156 29

36.5 56.9 10.7

36 43 6

42.4 50.5 7.1

63 64 54 50 43

23.0 23.4 19.7 18.2 15.7

8 8 15 23 31

9.4 9.4 17.6 27.1 36.5

39 100 134

14.5 36.6 48.9

4 20 60

4.8 23.6 70.6

0.59

0.30

0.21

<0.0001

0.0068

data.

they worked or the kind of work in which they could engage. As shown in Table 3, patients’ evaluation of the impact of other potential barriers to employment were not as clearly linked to their current employment status. That is, whether currently employed or not, a large percentage of all patients regarded the potential loss of disability benefits and access to health insurance as problematic. In an “exploratory” component of this study, patients were asked whether they thought their families and friends, physicians, social workers, nurses, technicians, and dietitians expected that dialysis patients should be employed. Their perceptions of these others’ beliefs with regard to their employment were significantly related to their actual employment status. As is depicted in

Table 3, in each case, the perception that others believed employment to be appropriate for dialysis patients was associated with a greater likelihood of employment. Logistic Modeling Table 4 reports the results of the multivariable analysis. The demographic variables included were patients’ educational level and predialysis employment status. The health and functioning variable included was the MOS total score. Both the MOS and Karnofsky scores were significantly associated with employment in the chisquare analyses. However, when they were included together in the multivariable analysis, the Kamofsky variable dropped out (probably due to shared variance). The attitudes and beliefs variables included

EMPLOYED

v NONEMPLOYED

DIALYSIS Table

PATIENTS

3. Patients’

Attitudes

537 and

Perceptions

(n = 359)

Nonemployed (n = 274)

Health-related limits Health limits no. of hours worked Yes No Health limits kind of work Yes No Other barriers Loss of disability benefits Yes No Access to health insurance Yes No Who believes patients can work Health team as whole Yes No Doctors Yes No Social workers Yes No Nurses Yes No Technicians Yes No Dietitians Yes No Family/friends Yes No Self (respondent) Yes No Half can work *Total

n may

vary

due to missing

Employed (n = 85)

n’

%

n*

%

213 39

84.5 15.5

39 43

47.6 52.4

215 37

85.3 14.7

51 34

60.0 40.0

210 56

78.9 21.1

64 19

77.1 22.9

208 56

78.8 21.2

65 18

78.4 21.7

104 155

40.2 59.8

54 28

65.9 34.1

159 102

60.9 39.1

65 17

79.3 20.7

57 97

61.8 38.2

67 15

81.7 18.3

114 111

46.5 43.5

61 20

75.3 24.7

134 115

53.8 46.2

57 22

72.1 27.8

139 110

55.8 44.2

62 18

77.6 22.5

75 191

28.2 71.8

47 35

57.3 42.7

26 126 116

9.7 47.0 43.3

18 22 43

21.7 26.5 51.8

Probability

Values

0.0001

0.0001

0.46

0.63

<0.0001

0.0010

0.0087

0.0071

0.0033

0.0016

<0.0001

0.0006

data.

were perceptions of limitations in ability to work (indexed score for limited in kind or amount of work able to do) and perceptions of others’ expectations regarding ability to work (indexed score for health team as a whole, doctors, social workers, nurses, technicians, dietitians, family/ friends, and self). As shown in Table 4, statistically significant

relationships were demonstrated between current employment status and predialysis employment status, education, MOS score, perceptions of limitations on working capability, and perceptions of others’ expectations regarding ability to work. A chi-square of 89.54 (P < 0.0001) resulted from this combination of variables (Table 4).

538

CURTIN Table

4. Logistic

Regression/Employment

Status

(Working)

on Significant

Independent B* 2 SE

Health and functioning variables MOS-SF 20 (total score) Economic factors Education Employed before dialysis Attitudes and beliefs Perceived limitations on ability to work (kind and amount) Perceptions of others’ beliefs/expectations regarding ability to work Model chi-square * Logistic

regression

Independent

ET AL

Variables

Variables Odds

Ratio

Probability

0.01

? 0.00

1.00

0.51 1.76

+ 0.13 ‘-’ 0.49

1.66 5.81

? 0.22

0.43

0.0001

0.07 i 0.03 89.54

1.07

0.0406
-0.84

Value

0.0048 <0.0001 <0.0001

coefficient.

DISCUSSION

Patients’ educational levels and their predialysis employment status both emerged as strong independent predictors of employment for patients on dialysis treatments.“32’-2” Explanations for these findings seem fairly straightforward, ie, more highly educated and/or “white-collar” workers are apt to have more job opportunities, to receive salaries that exceed disability incomes, to have insurance benefits available to them, to have jobs that require less heavy physical labor, and to have jobs that lend themselves to flexible work hours for accommodating dialysis scheduling. 83’1~24 Thus, for well-educated patients with a good work history, continuing employment (or re-employment) is quite feasible.9~2’~22Unfortunately, just the opposite set of circumstances works against the less-educated and/or previously unemployed dialysis patients.2’ Past findings with regard to the relationship between length of time on dialysis and employment have been inconsistent.8324 In this study, no differences were found between employed and unemployed patients with regard to length of time on dialysis or cause of renal failure, despite the fact that 25% of the patients were diabetic. Since this is consistent with’ but also contrary to findings in other research,239diagnostic categories were collapsed into “diabetic” versus “other” for further analysis and confirmation. Again, no statistically significant differences were observed. Although the objective health indicators did

not predict the work status of dialysis patients in this study, the MOS-SF 20 was significantly associated with employment. This finding confirms previous reports that have found the MOS, as well as other measures of overall functioning, to correlate positively with employment.9V25 Since these measures of functioning tend to be multidimensional and to assess physical functioning, role functioning, social functioning, mental health, health perceptions, and pain, they are able to portray patients’ overall sense of well-being, subjective health status, and ability to cope with their health situation. Thus, while objective indicators assess the numbers and severity of patients’ physical health problems, self-reported measures serve as indicators of how well patients are actually adapting and accommodating to their lives with renal failure. Significant differences exist between the attitudes of employed versus nonemployed patients with regard to health-related barriers to employment. For instance, employed patients reported that they were not limited by their health in the hours they worked or the kind of work in which they could engage.2’ However, since there was no significant difference found between working and nonworking subjects with regard to the objective health indicators, the substantial differences in attitude cannot be clearly attributed solely to a health status differential. Instead, patients who were not employed seemed to “perceive” their health as precarious and a serious problem vis-a-vis employment, even when the

EMPLOYED

v NONEMPLOYED

DIALYSIS

PATIENTS

objective characteristics of their health status did not confirm this.” In the past, many hypotheses concerning the impact of attitudes on rehabilitation outcomes for dialysis patients have been advanced. In this study, we explored the origin/nature of patients’ own expectations about their ability to work. This was accomplished by including a series of items asking subjects what they believed other people’s expectations are regarding dialysis patients’ ability to work. Employment status was found to relate in a significantly positive manner to patients’ perceptions of others’ expectations regarding their ability to work.” Since it has been shown that people, in general, perceive support for and consensus regarding their own beliefs (unless strong evidence to the contrary is presentedz6), this finding may not seem surprising. On the other hand, it might also suggest that dialysis patients may be taking their cues with regard to their potential for employment from those around them, a notion that has been suggested in previous research.“324 If this proves to be the case, dialysis patients’ vocational rehabilitation might be improved if providers are themselves convinced that ESRD and employment are compatible and if providers also are able to convince patients as we11.24 With the availability of meaningful predictors of dialysis patient employment, the prospect of successful vocational rehabilitation seems within reach. However, it would be useful to estimate the number of dialysis patients capable of and interested in returning to work.’ To this end, patients in this study were asked if they felt they were able to work and if they wanted to return to work. Of the 270 patients who were not currently employed, 66 (24%) reported that they were able to work full or part time and 182 (67%) said they would like to return to work full or part time. Twenty-one percent of the currently unemployed patients indicated that they would both like to return to work and were able to work either full or part time. Since only 24% of our sample are currently employed either full or part time, it is clear that bringing this additional 21% of patients to employment would constitute a significant increase in employed dialysis patients. The potential to increase the number of dialysis patients who are employed exists if the barriers to employment that have been identified can

539

be removed. Directing rehabilitation efforts toward the population of dialysis patients who identify themselves as willing and able to work seems a worthwhile strategy. The potential “payoff’ ’ in terms of dollars as well as human capital is likely to outweigh “costs” associated with such efforts. More importantly, the current poor rehabilitation outcomes reported among patients with ESRD would definitely be improved. ACKNOWLEDGMENT The authors acknowledge the invaluable contributions made by Cathy Mann and Jane Banning, without whose hard work and commitment to excellence this research would have been many times more difficult.

REFERENCES 1. Ifudu 0, Mayers J, Matthew J, Caridad C, Cambridge A, Friedman E: Dismal rehabilitation in geriatric inner-city hemodialysis patients. JAMA 271:29-33, 1994 2. Ifudu 0, Paul H, Mayers J, Cohen L, Brezsnyak A, Herman A, Friedman E: Pervasive failed rehabilitation in center-based maintenance hemodialysis patients. Am J Kidney Dis 23:394-400, 1994 3. Kutner N, Cardenas D, Bower J: Rehabilitation, aging and chronic renal disease. Am J Phys Med Rehabil 71:97101, 1992 4. Gordon D: As ESRD population grows older, difficult questions arise. Dial Transplant 18594.600, 1989 5. Blagg C, Fitts S: Rehabilitation and epoetin alfa in dialysis patients. New Dir Anemia 2:1-4, 1991 6. Evans R: Recombinant human erythropoietin and the quality of life of end-stage renal disease patients: A comparative analysis. Am J Kidney Dis 18:62-70, 1991 7. Levin N: Quality of life and hematocrit level. Am J Kidney Dis 20: 16-20, 1992 8. Holley J, Nespor S: An analysis of factors affecting employment of chronic dialysis patients. Am J Kidney Dis 23:681-685, 1994 9. Kutner N, Brogan D, Fielding B: Employment status and ability to work among working-age chronic dialysis patients. Am J Nephrol 11:334-340, 1991 10. Vachon RA: Rehabilitation: Defining its role in ESRD. Nephrol News Issues July:25 29-30, 45, 1992 11. Antonoff A, Mallinger M: Vocational rehabilitation: Limitations and resistance of renal patients. Dial Transplant 20:604-609, 1991 12. Domoto D: Rehabilitation of the end-stage renal disease patient: Are the right questions being asked? Am J Kidney Dis 23:467-468, 1994 13. USRDS, Annual Data Report: EPO and quality of life: Special Study 4, Appendix D, 1992 14. Carmines EG, Zeller RA: Validity, in Sullivan JL, Niemi RG (eds): Reliability and Validity Assessment. Beverly Hills, CA, Sage, 1988, pp 17-27 15. Nunnally J: Assessment of reliability, in Psychometric Theory. New York, NY, McGraw-Hill, 1978, pp 225-255 16. Stewart A, Hayes R, Ware J: The MOS short form general health survey. Med Care 25:724-735, 1988

540 17. Kamofsky D, Burchenal J: The clinical evaluation of chemotherapeutic agents in cancer, in McLeod CM (ed): Evaluation of Chemotherapeutic Agents. New York, NY, Columbia University Press, 1949, pp 191-205 18. Carlson D, Johnson W, Kjellstrand C: Functional status of patients with end-stage renal disease. Mayo Clin Proc 62:338-344, 1987 19. Harris L, Luft F, Rudy D, Tiemey W: Clinical correlates of functional status in patients with chronic renal insufficiency. Am J Kidney Dis 21:161-165, 1993 20. Kutner N, Cardenas D: Rehabilitation status of chronic renal disease patients undergoing dialysis: Variations by age category. Arch Phys Med Rehabil62:626-630, 1981 21. Femurs C, Powers M: The employment potential of hemodialysis patients. Nurs Res 34:273-277, 1985

CURTIN

ET AL

22. Florida ESRD Network Nineteen: Rehabilitation among Florida in-center dialysis patients. Unpublished report, 1981 23. Friedman N, Rogers T: Dialysis and the world of work. Contemp Dial Nephrol 9:16-19, 52, 1988 24. Rasgon S, Schwankovsky L, James-Rogers A, Wodrow L, Glick J, Butts E: An intervention for employment maintenance among blue-collar workers with end-stage renal disease. Am J Kidney Dis 22:403-412, 1993 25. Simmons R, Abress L: Quality-of-life issues for endstage renal disease patients. Am J Kidney Dis 15201-208, 1990 26. Backman CW: Attraction in interpersonal relationships, in Rosenberg M, Turner R (eds): Social Psychology: Sociological Perspectives. New York, NY, Basic Books, 1981, pp 235-268