- Email: [email protected]
Treatment Bias in the Management of End-Stage Renal Disease
ORIGINAL INVESTIGATIONS
Treatment Bias in the Management of End-Stage Renal Disease Marc D. Smith, PhD, Barry A. Hong, PhD, Jeffrey E. Michelman, MHA, and Alan M. Robson, MD, FRCP • A study was conducted of 419 patients with end-stage renal disease (ESRD) being treated by center or home hemodialysis or by renal transplantation at four facilities located within 2.5 km of each other. The objectives were to examine the distribution of patients among the three modes of treatment and to analyze patient transfers to alternate modes of ESRD therapy. While white patients at each facility were comparable (P > 0.05) on age, sex, travel time to treatment, marital status, work or employment status, and the presence of diabetes mellitus, the distribution of patients among the treatment modes differed significantly (P < 0.001) across the facilities. Similarly, the sociodemographic and diagnostic characteristics of the nonwhite patients were comparable at each of the facilities (P > 0.05); however, despite observable variation among the facilities in the distribution of these patients, the differences did not achieve statistical significance (P > 0.05). Patient transfers to alternate modes of ESRD therapy were infrequent, and among center hemodialysis patients, the distribution of transfers differed significantly across the facilities (P < 0.001). It is concluded that the distribution of patients was dependent on the patient's initial mode of therapy and the staff attitudes at the individual facilities. INDEX WORDS: ESRD; hemodialysis; transplantation.
S
INCE THE ENACTMENT of Public Law 92603 (1972) extending Medicare coverage to virtually all patients with end-stage renal disease (ESRD), attention has increasingly focused on the rapidly escalating economic costs of the program, 13 the markedly increased number of participating patients,46 and the distribution of patients among the various modes of ESRD therapy.79 Once operational, the program initially provided coverage for 23,000 maintenance dialysis and transplant patients at an annual cost of $283 million. 6 By 1978, however, the number of participating patients and the annual federal subsidy had risen to 50,000 and $947 million, respectively. Should the current patterns of treatment selection and utilization continue, it has been estimated that by 1995, 90,000 patients will be enrolled in the program at an annual cost approaching $4.6 billion. 6 Concerns regarding the total cost of treatment and the scope of patient participation are not, therefore, without foundation. Recognizing the current concentration of ESRD patients in center hemodialysis 810 and the economic consequences associated with such a distribution, recent investigations 79 have examined the cost-effectiveness of alternate therapies, particularly home hemodialysis and renal transplantation. Stange and Sumner7 concluded that the transition of 1,000 patients from center hemodialysis to transplantation in each of 10 successive years would be accompanied by a cost reduction of $279
American Journal of Kidney Diseases, Vol III, No 1, July 1983
to $330 million. More modest decreases in cost could be anticipated in the similar transition of patients from home hemoaialysis to transplantation ($103 to $142 million) and from center to home hemodialysis ($241 million). Similarly, the results of the simulation model developed by Roberts et al 9 suggested potential cost reductions of an equivalent magnitude. Marked differences in the levels of psychosocial adaptation and rehabilitation have also been observed with various modes of ESRD therapy. 11 17 Although advantages and limitations are inherent within each therapeutic mode, high levels of adaptation and rehabilitation have been more typically cited for home hemodialysis and transplantation than for center hemodialysis.II.18.19 Blagg et aP2 and Drees and Gallagher l l have observed, for example, a greater sense of personal autonomy and control among home hemodialysis patients than From the Health Administration and Planning Program and the Departments of Psychiatry and Pediatrics, Washington University School of Medicine, St Louis, Mo; The Jewish Hospital of St Louis, St Louis, Mo; and Division of Pediatric Nephrology, St Louis Children's Hospital, St Louis, Mo. Address reprint requests to Marc D. Smith, PhD, Assistant Professor, Health Administration and Planning Program. Washington University School of Medicine, 4547 Clayton Avenue, St Louis, MO 63110. This research was supported by a grant from the Missouri Kidney Program. © 1983 by The National Kidney Foundation, Inc. 0272-6386/83/01021-06$01.00/0
21
22
among center patients. Similarly, improved work adjustment 20 and health status 21 have been reported from home hemodialysis patients. When contrasted with both center and home hemodialysis, transplantation has been associated with a more frequent return to pre morbid levels of activity,22 improved physical capacity,23 and decreased episodes of illness and hospitalization. 7 Although the economic and psychosocial incentives associated with both home hemodialysis and transplantation have been frequently reported in the literature, the experience of the ESRD program indicates that center hemodialysis continues to be the preferred mode of therapy. 8 IO Several characteristics of both the ESRD patient population and service delivery system that appear to have contributed to this trend are as follows: the increasing age of the patient population 4 and prevalence of systemic disease;24 the static availability of donor kidneysIO and the decreased survival of organ grafts6 (athough recent innovations, including pretransplant blood transfusions, suggest the potential for a reversal of this trend); 25 the absence of financial incentives for home hemodialysis;46 and a decline in the socioeconomic and educational status of ESRD patients. 4 Acknowledging the disproportionate distribution of patients on center hemodialysis, several investigators 9 . I726 have suggested that the initiation of center therapy may predispose patients to this mode throughout the course of their illness. While the empirical evidence is scarce, it would appear that the process of patient adaptation to center hemodialysis is of sufficient intensity to preclude transfer to an alternate therapeutic mode (home hemodialysis or transplantation). 26 To examine the validity of this observation, as well as the extent to which the process of adaptation is uniformly applicable to ESRD patients irrespective of their mode of therapy, a retrospective panel study was conducted at four metropolitan dialysis and transplant facilities. Consistent with the national and international data previously reported by ReIman and Rennie,8 it was hypothesized that the distribution of patients among ESRD treatment modes was dependent on factors unrelated to the characteristics of the patients themselves. Similarly, it was believed that the modes of maintenance therapy to which patients transferred were also independent of the inherent characteristics of the patients.
SMITH ET AL
MATERIALS AND METHODS A questionnaire was developed by the investigators to identify adult* patients who had received continuous treatment for chronic renal failure from August I, 1979 to July 31, 1980 and to determine the frequency of patient transfers among modes of maintenance therapy during the observation period. In addition, the survey instrument was designed to determine the extent to which therapeutic modes were dependent on the sociodemographic or diagnostic characteristics of the patients, the patients' mode of ESRD therapy at the beginning of the study, or the treatment facility. The modes of maintenance treatment considered were center hemodialysis, home hemodialysis, renal transplantation, and the various combinations of these therapies that occurred within the sample. Neither continuous ambulatory nor intermittent peritoneal dialysis were available for patients included in the study. The classification of patients by mode of treatment at the beginning of the observation period reflected their usual mode of therapy. Temporary use of a therapy to remedy a transient medical or psychosocial problem was not considered in the analysis since long-term modes of therapy were unaffected by such intervention. Thus, for example, a home hemodialysis patient who was receiving center therapy to stabilize a medical complication prior to a return to the home setting was classified as a home hemodialysis patient. The questionnaire was forwarded to the clinical directors of four dialysis and transplant facilities in a midwestern health service area; in excess of 90% of the 2.3 million residents of this predominantly urban area live within one hour travel time of these treatment facilities 27 The remaining two facilities located within the community were excluded from the survey; one had experienced a significant interruption in services, while the other primarily served a population external to the health service area. Responses were obtained from each of the four facilities and were subsequently verified by the investigators through a personal review of individual patient records. Patients were classified by treatment facility, mode of maintenance therapy, age, race, sex, travel time to the treatment facility (less than 30 minutes versus greater than 30 minutes,
one-way), marital status, work or employment status,28 and the presence or absence of diabetes mellitus. Appropriate descriptive measures have been employed to report the frequency data. The analysis of categorical data has been accomplished through the application of the X2 test for statistical independence, while the comparability of patient ages among the facilities was evaluated by a one-way analysis of variance. 29 Prior to the acquisition of data, the facilities were informed of the research objectives and the confidentiality of their responses. They were also assured that the data would be analyzed and reported in a manner precluding the identification of either the individual facilities or their patients.
RESULTS
Data were obtained on 419 maintenance ESRD patients. This represented 100% of the patients *The 33 pediatric patients treated for ESRD within the health service area during the observation period received care at a separate pediatric facility not included in the study.
23
TREATMENT BIAS IN ESRD Table 1.
Characteristics of the ESRD Facilities Patients (n)
Facility
A B C D Total
Location
Transplant Service"
1,2 2
Hospital Hospital Hospital Free standing
Typet
Nephrologist Compensation:j:
Center Hemodialysis
Home Hemodialysis
Trans· plantation
Total
NFP NFP NFP FP
FTS FTS PTS/FFS FTS
87 25 20 70 202
26 17 3 9 55
109 14 15 24 162
222 56 38 103 419
* 1 = University service affiliated with facilities A and B; 2 tNFP = not for profit; FP = for profit. :j:FTS = full-time salaried; PTS = part-time salaried; FFS
who fulfilled the criteria for inclusion in the study at the four participating facilities. Since two additional facilities were not surveyed (see Materials and Methods), the patient sample studied represented 85.1 % of the patients eligible for participation within the entire health service area. A description of the facilities is presented in Table 1. The median age of the 419 patients was 45.6 years; 51.1 % were male and 64.2 % were white (Table 2). Neither the ages of the patients nor their distribution by sex, travel time to treatment, marital status, work or employment status, or the presTable 2.
=
university service affiliated with VA and facility D.
=
fee-for-service.
ence of diabetes mellitus differed significantly among the four facilities (P > 0.05). A statistically significant difference was observed, however, in the racial composition of the facilities (P < 0.05). Since the data demonstrated the distribution of patients among the three modes of therapy to be race-dependent (P < 0.001), further analyses were performed holding race constant. Among the 269 white patients receiving maintenance ESRD therapy at the beginning of the observation period (Table 2), neither their ages nor their distribution by sex, travel time to treatment,
Patient Characteristics and Treatment Modes at Initiation of Study
Facility
A
B
C
D
All White Patients
No. of Patients Patient characteristics Median age* Sext (% male) Travel time:j: Diabetes mellitus (%) Marital status§ Single Married Separated Divorced Widowed Work/employment status§ Unemployed or never worked Unskilled or semiskilled Skilled or white-collar Part-time housework Full-time housework Treatment Mode Center hemodialysis§ Home hemodialysis§ Transplantation§
149
38
28
54
269
73
18
10
49
150
419
38.9 49.7 93.3 10.1
48.5 47.4 84.4 5.3
43.8 53.6 78.6 17.9
48.2 57.4 90.6 11.1
46.0 51.3 89.9 10.4
48.1 43.8 39.7 8.2
46.8 44.4 61.1 5.6
50.5 70.0 22.2 10.0
49.1 59.2 36.4 8.9
48.0 50.7 40.3 8.0
45.6 51.1 72.1 9.5
24.2 68.4 0 3.4 4.0
23.7 65.8 2.6 5.3 2.6
28.6 57.1 0 10.7 3.6
16.7 72.2 0 7.4 3.7
23.0 67.7 0.4 5.2 3.7
24.7 56.2 5.5 6.8 6.8
16.7 66.7 11.1 0 5.5
30.0 50.0 0 20.0 0
30.6 61.2 2.0 4.1 2.0
26.0 58.7 4.7 6.0 4.7
24.1 64.4 1.9 5.5 4.1
19.5 22.8 37.6 0 20.1
26.3 18.4 31.6 0 23.7
10.7 17.9 42.8 0 28.6
18.5 27.8 40.7 0 13.0
19.3 22.7 37.9 0 20.1
28.8 39.7 17.8 0 13.7
33.3 50.0 5.6 5.6 5.6
50.0 20.0 0 10.0 20.0
28.6 40.8 24.5 0 6.1
30.7 40.0 17.3 1.3 10.7
23.4 28.9 30.5 0.5 16.7
23.5 14.8 61.7
36.8 34.2 28.9
46.4 10.7 42.9
61.1 14.8 24.1
35.3 17.1 47.6
71.2 5.5 23.3
61.1 22.2 16.7
70.0 0 30.0
75.5 2.0 22.4
71.3 6.0 22.7
48.2 12.9 38.9
White Patients
*Expressed tExpressed :j:Expressed §Expressed
in years. as percentage of males. as percentage greater than 30 minutes. in percent.
D
All Nonwhite Patients
All Patients
Nonwhite Patients A
B
C
24
SMITH ET AL
marital status, work or employment status, or the presence of diabetes mellitus differed significantly across the four facilities (P > 0.05). Ninety-five (35.3 %) of these patients were receiving center hemodialysis, 46 (17.1 %) were on home hemodialysis, and 128 (47.6%) had functioning kidney transplants. The distribution of patients among the three therapeutic modes differed significantly across the facilities (P < 0.001). Thus, for example, the percentage of transplant recipients was 2.6 times greater at facility A than at facility D. Furthermore, the distribution of patients between center and noncenter (the sum of home hemodialysis and transplantation) therapies also differed significantly across the facilities (P < 0.001). Among white patients, therefore, the mode of therapy was dependent on the facility at which the patient received treatment. At the beginning of the study, 150 nonwhite patients were receiving maintenance therapy for ESRD (Table 2). Neither the ages of the patients nor their distribution by sex, travel time to treatment, marital status, work or employment status, or the presence of diabetes mellitus differed significantly across the four facilities (P > 0.05). Among the nonwhite patients, 107 (71. 3 %) were receivng center hemodialysis, 9 (6.0%) were on home hemodialysis, and 34 (22.7%) had functioning kidney transplants. Neither the distribution of patients among the three therapeutic modes nor the distribution of patients between center and noncenter therapies differed significantly across the facilities (P > 0.05). Thus, while obvious differences in the frequency of nonwhite patients receiving various modes of therapy were observed among the facilities, they were not statistically significant. In addition to the analysis of patient distribution among modes of maintenance therapy at the beginning of the study, the transfer of patients to alternate therapeutic modes during the observation peTable 3.
riod was also examined. Of the 202 patients receiving center hemodialysis at the initiation of the study, only 28 (13.9%) transferred to an alternate therapeutic mode (Table 3); the preponderance of those who transferred received a renal transplant (82.1 %). Furthermore, the mode of treatment to which center hemodialysis patients transferred differed significantly across the four facilities (P < 0.001). Thus, for example, the frequency of transfers to transplantation among center hemodialysis patients at facility C was 3.5 times greater than at facility A. Similarly, the frequency of transfers to home hemodialysis at facility B was 7.3 times greater than the frequency at facility A and 3.2 times greater than the mean for the four facilities. Since the ages of the center hemodialysis patients and their distribution by race, sex, travel time to treatment, marital status, work or employment status, and the presence of diabetes mellitus did not differ significantly among the facilities (P > 0.05), the type of transfer among center hemodialysis patients was dependent on the facility at which they were treated. Among the 55 patients receiving home hemodialysis at the beginning of the study, 11 (20.0%) transferred to an alternate therapy (Table 3). Similar to center hemodialysis patients, the majority of home patients who transferred received a renal transplant (72.7%). Of the 162 transplant recipients at the initiation of the study, 11 (6.8 %) rejected the graft during the observation period and required a return to hemodialysis. Among the 11 patients, however, only two did not pursue further renal transplantation. The insufficient size of these samples precluded a specific analysis of the potential facility dependence of transfers from home hemodialysis and transplantation. DISCUSSION
In their examination of dialysis and transplant rates, ReIman and Rennie 8 observed a substantial
ESRD Patients Transferring to Alternate Treatment Modes by Facility and Therapeutic Mode' CH Patients Transferring to
T Patients Transferring to
HH Patients Transferring to
Facility
HH
T
Total
CH
T
Total
CH
HH
Total
A
1.1 8.0 5.0 1.4 2.5
5.7 12.0 20.0 15.7 11.4
6.9 20.0 25.0 17.1 13.9
3.8 5.9 0 11.1 5.5
23.1 11.8 0 0 14.5
26.9 17.6 0 11.1 20.0
0 0 0 4.2 0.6
0 7.1 0 0 0.6
0 7.1 0 4.2 1.2
B C D Total
CH = center hemodialysis; HH * All numbers are percentages.
=
home hemodialysis; T
= transplantation.
25
TREATMENT BIAS IN ESRD
variation within both the United States and the international community. Similarly, data developed in the present study demonstrated a significant difference in the distribution of patients across ESRD treatment modes among individual facilities within the same health service area. While patients at .the four participating facilities were comparable on age, sex, travel time to treatment, marital status, work or employment status , and the presence of diabetes mellitus , a statistically significant difference in the racial composition of the facilities was also apparent. A more detailed analysis of the data revealed the mode of maintenance therapy to be facility-dependent among the white ESRD patients who comprised 64 .2 % of the patients studied. For example, among these patients the percentage receiving center hemodialysis at facility D was 2.6 times greater than at facility A (Table 2); the percentage of home hemodialysis patients was 3.2 times greater at facility B than at facility C. While a statistically significant difference in the frequency of nonwhite patients receiving various modes of ESRD therapy was not present across the facilities, substantial variation was nonetheless observed; for example, 22.2 % of the nonwhite patients at facility B were on home hemodialysis compared with 0 % at facility C. Although sociodemographic and diagnostic differences among the white patients might be proposed as an explanation,30 the data clearly did not support this argument. White patients treated at the four facilities were comparable on age, sex, travel time to treatment, marital status, work or employment status, and the presence of diabetes mellitus. This was not surprising since the facilities are located within 2.5 km of each other. It might also be hypothesized that home hemodialysis and renal transplantation were perceived as competitive therapies; thus , the selection of either therapeutic mode would merely reflect the physician or patient preference for a therapy of equivalent benefit. However, the statistically significant difference in the distribution of white patients between center and noncenter therapies across the facilities precluded this explanation. It would appear, therefore , that the bias observed in the treatment of these patients reflected the attitudes of the staff and patients toward the three modes of therapy at the respective facilities , rather than inherent sociodemographic or diagnostic differences in the patient populations served.
While the retrospective design of the study did not permit the direct examination of the importance of these attitudes, additional support for this conclusion was evidenced by the facility dependence of patient transfers from center hemodialysis to alternate modes of maintenance therapy. It is also noteworthy that the available evidence did not suggest that the treatment differences observed in this study were a consequence of the attitudes of referring physicians or their knowledge of the therapeutic preferences of individual facilities. Irrespective of the distribution of patients among treatment modes and the substantial progress evidenced by several facilities in achieving comparatively high rates of transfer to home hemodialysis and transplantation, it was equally apparent that patients continued their current mode of therapy. The stability of patients within each of the therapies suggested that adaptation 31 ' 36 was associated with specific therapeutic modes , rather than with ESRD therapy in general. The implications of these findings for both the formulation of public policy and the management of patient care are substantial. The probable relationship between staff attitudes and the selection of treatment modes suggests that further attention be directed toward improved professional education as well as peer support and review. Furthermore, the intensity of adaptation to specific modes of maintenance therapy implies that if home hemodialysis and transplantation are to be substantive components of a national ESRD policy, they should be employed, with increased frequency, as the initial modes of treatment. ACKNOWLEDGMENT The authors wish to express their appreciation to David J. Baltzer and Thomas J. Valerius for their invaluable assistance in the collection of the data presented in this study.
REFERENCES I. Hoffstein PA, Krueger KK , Wineman RJ: Dialysis costs: Results of a diverse sample study. Kidney Int 9:286-293, 1976 2. Burton BT: End-stage renal disease: Better treatment at lower cost. Ann Intern Med 88:567-568 , 1978 3. Adams LR: Medicare coverage for chronic renal disease: Policy implications. Health Soc Work 3:41-53, 1978 4 . Evans RW, Blagg CR , Bryan FA: Implications for health care policy: A social and demographic profile of hemodialysis patients in the United States. JAMA 245 :487-491 , 1981 5. Guttmann RD: Renal transplantation. N Engl J Med 301 :975-982 , 1979
26 6. Rettig RA: The politics of health cost containment: Endstage renal disease. Bull NY Acad Med 56:115-137, 1980 7. Stange PV, Sumner AT: Predicting treatment costs and life expectancy for end-stage renal disease. N Engl J Med 298:372-378, 1978 8. Reiman AS, Rennie D: Treatment of end-stage renal disease: Free but not equal. N Engl J Med 303:996-998, 1980 9. Roberts SD, Maxwell DR, Gross TL: Cost-effective care of end-stage renal disease: A billion dollar question. Ann Intern Med 92:243-248, 1980 10. Wineman RJ: End-stage renal disease: 1978. Dial Transplant 7:1034-1065,1978 11.Drees A, Gallagher EB: Hemodialysis, rehabilitation, and psychological support, in Levy NB (ed): Psychological Factors in Hemodialysis and Transplantation. New York, Plenum, 1981, pp 133-145 12. Blagg CR, Hickman RO, Eschbach JW, et al: Home hemodialysis: Six years' experience. N Engl J Med 283: 11261131, 1970 13. Cadnapaphornchai P, Kuruvila KC, Holmes J, et al: Analysis of five years' experience of home dialysis as a treatment modality for patients with end-stage renal failure. Am J Med 57:789-799, 1974 14. Roberts JL: Analysis and outcome of 1,068 patients trained for home hemodialysis. Kidney Int 9:363-374, 1976 15. Salvatierra 0: Renal transplantation in perspective. Dial Transplant 7:171-180,1978 16. Havry P: Dialysis and kidney transplantation in Europe. Med Bioi 55:245-248, 1977 17. Whalen JE, Freeman RM: Home hemodialysis review in Iowa. Arch Intern Med 138:1787-1790, 1978 18. Brenner BM, Lazarus JM: Chronic renal failure, in Isselbacher KJ, Adams RD, Braunwald E, et a1 (eds): Harrison's Principles ofInternal Medicine. St Louis, McGraw-Hill, 1980, pp 1299-1307 19. Blagg CR, Scribner BH: Dialysis: Medical, psychosocial, and economic problems unique to the dialysis patient, in Brenner BM, Rector FC (eds): The Kidney. Philadelphia, Saunders, 1976, pp 1705-1744 20. Robinson BHB: Intermittent haemodialysis in the home. Br Med Bull 27: 173-180, 1971
SMITH ET AL 21. Mock LA, Kopel K: Psychosocial aspects of home and in-center dialysis. Dial Transplant 6:40-43, 1977 22. Levine C: Dialysis or transplant: Values and choices. Hastings Cent Rep 8:8-10, 1978 23. Dutz H, Mebel M, Roseler E, et al: Evaluation of factors involved in rehabilitation using dialysis and transplantation. Proc 6th Int Congr Nephrol. Basel, Switzerland, Karger, 1976, pp 710-714 24. Anonymous: Selection of patients for dialysis and transplantation. Br Med J 2:1449-1450, 1978 25. Guttmann RD: Renal transplantation. N Engl J Med 301:1038-1048,1979 26. Levy NB, Abram HS, Kemph Jp, et al: Panel: Living or dying: Adaptation to hemodialysis, in Levy NB (ed): Living or Dying: Adaptation to Hemodialysis. Springfield, Ill, Charles C Thomas, 1974, pp 3-29 27. Greater St Louis Health Systems Plan: 1980-1984. St Louis, Greater St LOUIS Health Systems Agency, 1980, pp ESRD 1-13 28. Gutman RA, Stead WW, Robinson RR: Physical activity and employment status of patients on maintenance dialysis. N Engl J Med 304:309-313, 1981 29. Nie NH, Hull CH , Jenkins JG, et al: SPSS: Statistical Package for the Social Sciences (ed 2). New York, McGrawHill,1975 30. Alexandre GPJ: Factors involved in choosing dialysis or transplantation. Proc 6th Int Congr Nephrol, Basel, Switzerland, Karger, 1976, pp 692-697 31. Seime RJ, Zimmerman J: Adjustment to dialysis: A brief review. Behav Med Update 3:9-11, 1981 32. De-Nour AK, Shaltiel J, Czaczkes JW: Emotional reactions of patients on chronic hemodialysis. Psychosom Med 30:521-533, 1968 33. Levy NB: The psychology and care of the maintenance hemodialysis patient. Heart Lung 2:400-405, 1973 34. Flannery JG: Adaptation to chronic renal failure. Psychosomatics 19:784-787, 1978 35. Abram HS: The psychiatrist, the treatment of chronic renal failure, and the prolongation of life: II. Am J Psychiatry 126:157-166, 1969 36. Levy NB: Psychological sequelae to hemodialysis. Psychosomatics 19:329-331, 1978
Treatment Bias in the Management of End-Stage Renal Disease Marc D. Smith, PhD, Barry A. Hong, PhD, Jeffrey E. Michelman, MHA, and Alan M. Robson, MD, FRCP • A study was conducted of 419 patients with end-stage renal disease (ESRD) being treated by center or home hemodialysis or by renal transplantation at four facilities located within 2.5 km of each other. The objectives were to examine the distribution of patients among the three modes of treatment and to analyze patient transfers to alternate modes of ESRD therapy. While white patients at each facility were comparable (P > 0.05) on age, sex, travel time to treatment, marital status, work or employment status, and the presence of diabetes mellitus, the distribution of patients among the treatment modes differed significantly (P < 0.001) across the facilities. Similarly, the sociodemographic and diagnostic characteristics of the nonwhite patients were comparable at each of the facilities (P > 0.05); however, despite observable variation among the facilities in the distribution of these patients, the differences did not achieve statistical significance (P > 0.05). Patient transfers to alternate modes of ESRD therapy were infrequent, and among center hemodialysis patients, the distribution of transfers differed significantly across the facilities (P < 0.001). It is concluded that the distribution of patients was dependent on the patient's initial mode of therapy and the staff attitudes at the individual facilities. INDEX WORDS: ESRD; hemodialysis; transplantation.
S
INCE THE ENACTMENT of Public Law 92603 (1972) extending Medicare coverage to virtually all patients with end-stage renal disease (ESRD), attention has increasingly focused on the rapidly escalating economic costs of the program, 13 the markedly increased number of participating patients,46 and the distribution of patients among the various modes of ESRD therapy.79 Once operational, the program initially provided coverage for 23,000 maintenance dialysis and transplant patients at an annual cost of $283 million. 6 By 1978, however, the number of participating patients and the annual federal subsidy had risen to 50,000 and $947 million, respectively. Should the current patterns of treatment selection and utilization continue, it has been estimated that by 1995, 90,000 patients will be enrolled in the program at an annual cost approaching $4.6 billion. 6 Concerns regarding the total cost of treatment and the scope of patient participation are not, therefore, without foundation. Recognizing the current concentration of ESRD patients in center hemodialysis 810 and the economic consequences associated with such a distribution, recent investigations 79 have examined the cost-effectiveness of alternate therapies, particularly home hemodialysis and renal transplantation. Stange and Sumner7 concluded that the transition of 1,000 patients from center hemodialysis to transplantation in each of 10 successive years would be accompanied by a cost reduction of $279
American Journal of Kidney Diseases, Vol III, No 1, July 1983
to $330 million. More modest decreases in cost could be anticipated in the similar transition of patients from home hemoaialysis to transplantation ($103 to $142 million) and from center to home hemodialysis ($241 million). Similarly, the results of the simulation model developed by Roberts et al 9 suggested potential cost reductions of an equivalent magnitude. Marked differences in the levels of psychosocial adaptation and rehabilitation have also been observed with various modes of ESRD therapy. 11 17 Although advantages and limitations are inherent within each therapeutic mode, high levels of adaptation and rehabilitation have been more typically cited for home hemodialysis and transplantation than for center hemodialysis.II.18.19 Blagg et aP2 and Drees and Gallagher l l have observed, for example, a greater sense of personal autonomy and control among home hemodialysis patients than From the Health Administration and Planning Program and the Departments of Psychiatry and Pediatrics, Washington University School of Medicine, St Louis, Mo; The Jewish Hospital of St Louis, St Louis, Mo; and Division of Pediatric Nephrology, St Louis Children's Hospital, St Louis, Mo. Address reprint requests to Marc D. Smith, PhD, Assistant Professor, Health Administration and Planning Program. Washington University School of Medicine, 4547 Clayton Avenue, St Louis, MO 63110. This research was supported by a grant from the Missouri Kidney Program. © 1983 by The National Kidney Foundation, Inc. 0272-6386/83/01021-06$01.00/0
21
22
among center patients. Similarly, improved work adjustment 20 and health status 21 have been reported from home hemodialysis patients. When contrasted with both center and home hemodialysis, transplantation has been associated with a more frequent return to pre morbid levels of activity,22 improved physical capacity,23 and decreased episodes of illness and hospitalization. 7 Although the economic and psychosocial incentives associated with both home hemodialysis and transplantation have been frequently reported in the literature, the experience of the ESRD program indicates that center hemodialysis continues to be the preferred mode of therapy. 8 IO Several characteristics of both the ESRD patient population and service delivery system that appear to have contributed to this trend are as follows: the increasing age of the patient population 4 and prevalence of systemic disease;24 the static availability of donor kidneysIO and the decreased survival of organ grafts6 (athough recent innovations, including pretransplant blood transfusions, suggest the potential for a reversal of this trend); 25 the absence of financial incentives for home hemodialysis;46 and a decline in the socioeconomic and educational status of ESRD patients. 4 Acknowledging the disproportionate distribution of patients on center hemodialysis, several investigators 9 . I726 have suggested that the initiation of center therapy may predispose patients to this mode throughout the course of their illness. While the empirical evidence is scarce, it would appear that the process of patient adaptation to center hemodialysis is of sufficient intensity to preclude transfer to an alternate therapeutic mode (home hemodialysis or transplantation). 26 To examine the validity of this observation, as well as the extent to which the process of adaptation is uniformly applicable to ESRD patients irrespective of their mode of therapy, a retrospective panel study was conducted at four metropolitan dialysis and transplant facilities. Consistent with the national and international data previously reported by ReIman and Rennie,8 it was hypothesized that the distribution of patients among ESRD treatment modes was dependent on factors unrelated to the characteristics of the patients themselves. Similarly, it was believed that the modes of maintenance therapy to which patients transferred were also independent of the inherent characteristics of the patients.
SMITH ET AL
MATERIALS AND METHODS A questionnaire was developed by the investigators to identify adult* patients who had received continuous treatment for chronic renal failure from August I, 1979 to July 31, 1980 and to determine the frequency of patient transfers among modes of maintenance therapy during the observation period. In addition, the survey instrument was designed to determine the extent to which therapeutic modes were dependent on the sociodemographic or diagnostic characteristics of the patients, the patients' mode of ESRD therapy at the beginning of the study, or the treatment facility. The modes of maintenance treatment considered were center hemodialysis, home hemodialysis, renal transplantation, and the various combinations of these therapies that occurred within the sample. Neither continuous ambulatory nor intermittent peritoneal dialysis were available for patients included in the study. The classification of patients by mode of treatment at the beginning of the observation period reflected their usual mode of therapy. Temporary use of a therapy to remedy a transient medical or psychosocial problem was not considered in the analysis since long-term modes of therapy were unaffected by such intervention. Thus, for example, a home hemodialysis patient who was receiving center therapy to stabilize a medical complication prior to a return to the home setting was classified as a home hemodialysis patient. The questionnaire was forwarded to the clinical directors of four dialysis and transplant facilities in a midwestern health service area; in excess of 90% of the 2.3 million residents of this predominantly urban area live within one hour travel time of these treatment facilities 27 The remaining two facilities located within the community were excluded from the survey; one had experienced a significant interruption in services, while the other primarily served a population external to the health service area. Responses were obtained from each of the four facilities and were subsequently verified by the investigators through a personal review of individual patient records. Patients were classified by treatment facility, mode of maintenance therapy, age, race, sex, travel time to the treatment facility (less than 30 minutes versus greater than 30 minutes,
one-way), marital status, work or employment status,28 and the presence or absence of diabetes mellitus. Appropriate descriptive measures have been employed to report the frequency data. The analysis of categorical data has been accomplished through the application of the X2 test for statistical independence, while the comparability of patient ages among the facilities was evaluated by a one-way analysis of variance. 29 Prior to the acquisition of data, the facilities were informed of the research objectives and the confidentiality of their responses. They were also assured that the data would be analyzed and reported in a manner precluding the identification of either the individual facilities or their patients.
RESULTS
Data were obtained on 419 maintenance ESRD patients. This represented 100% of the patients *The 33 pediatric patients treated for ESRD within the health service area during the observation period received care at a separate pediatric facility not included in the study.
23
TREATMENT BIAS IN ESRD Table 1.
Characteristics of the ESRD Facilities Patients (n)
Facility
A B C D Total
Location
Transplant Service"
1,2 2
Hospital Hospital Hospital Free standing
Typet
Nephrologist Compensation:j:
Center Hemodialysis
Home Hemodialysis
Trans· plantation
Total
NFP NFP NFP FP
FTS FTS PTS/FFS FTS
87 25 20 70 202
26 17 3 9 55
109 14 15 24 162
222 56 38 103 419
* 1 = University service affiliated with facilities A and B; 2 tNFP = not for profit; FP = for profit. :j:FTS = full-time salaried; PTS = part-time salaried; FFS
who fulfilled the criteria for inclusion in the study at the four participating facilities. Since two additional facilities were not surveyed (see Materials and Methods), the patient sample studied represented 85.1 % of the patients eligible for participation within the entire health service area. A description of the facilities is presented in Table 1. The median age of the 419 patients was 45.6 years; 51.1 % were male and 64.2 % were white (Table 2). Neither the ages of the patients nor their distribution by sex, travel time to treatment, marital status, work or employment status, or the presTable 2.
=
university service affiliated with VA and facility D.
=
fee-for-service.
ence of diabetes mellitus differed significantly among the four facilities (P > 0.05). A statistically significant difference was observed, however, in the racial composition of the facilities (P < 0.05). Since the data demonstrated the distribution of patients among the three modes of therapy to be race-dependent (P < 0.001), further analyses were performed holding race constant. Among the 269 white patients receiving maintenance ESRD therapy at the beginning of the observation period (Table 2), neither their ages nor their distribution by sex, travel time to treatment,
Patient Characteristics and Treatment Modes at Initiation of Study
Facility
A
B
C
D
All White Patients
No. of Patients Patient characteristics Median age* Sext (% male) Travel time:j: Diabetes mellitus (%) Marital status§ Single Married Separated Divorced Widowed Work/employment status§ Unemployed or never worked Unskilled or semiskilled Skilled or white-collar Part-time housework Full-time housework Treatment Mode Center hemodialysis§ Home hemodialysis§ Transplantation§
149
38
28
54
269
73
18
10
49
150
419
38.9 49.7 93.3 10.1
48.5 47.4 84.4 5.3
43.8 53.6 78.6 17.9
48.2 57.4 90.6 11.1
46.0 51.3 89.9 10.4
48.1 43.8 39.7 8.2
46.8 44.4 61.1 5.6
50.5 70.0 22.2 10.0
49.1 59.2 36.4 8.9
48.0 50.7 40.3 8.0
45.6 51.1 72.1 9.5
24.2 68.4 0 3.4 4.0
23.7 65.8 2.6 5.3 2.6
28.6 57.1 0 10.7 3.6
16.7 72.2 0 7.4 3.7
23.0 67.7 0.4 5.2 3.7
24.7 56.2 5.5 6.8 6.8
16.7 66.7 11.1 0 5.5
30.0 50.0 0 20.0 0
30.6 61.2 2.0 4.1 2.0
26.0 58.7 4.7 6.0 4.7
24.1 64.4 1.9 5.5 4.1
19.5 22.8 37.6 0 20.1
26.3 18.4 31.6 0 23.7
10.7 17.9 42.8 0 28.6
18.5 27.8 40.7 0 13.0
19.3 22.7 37.9 0 20.1
28.8 39.7 17.8 0 13.7
33.3 50.0 5.6 5.6 5.6
50.0 20.0 0 10.0 20.0
28.6 40.8 24.5 0 6.1
30.7 40.0 17.3 1.3 10.7
23.4 28.9 30.5 0.5 16.7
23.5 14.8 61.7
36.8 34.2 28.9
46.4 10.7 42.9
61.1 14.8 24.1
35.3 17.1 47.6
71.2 5.5 23.3
61.1 22.2 16.7
70.0 0 30.0
75.5 2.0 22.4
71.3 6.0 22.7
48.2 12.9 38.9
White Patients
*Expressed tExpressed :j:Expressed §Expressed
in years. as percentage of males. as percentage greater than 30 minutes. in percent.
D
All Nonwhite Patients
All Patients
Nonwhite Patients A
B
C
24
SMITH ET AL
marital status, work or employment status, or the presence of diabetes mellitus differed significantly across the four facilities (P > 0.05). Ninety-five (35.3 %) of these patients were receiving center hemodialysis, 46 (17.1 %) were on home hemodialysis, and 128 (47.6%) had functioning kidney transplants. The distribution of patients among the three therapeutic modes differed significantly across the facilities (P < 0.001). Thus, for example, the percentage of transplant recipients was 2.6 times greater at facility A than at facility D. Furthermore, the distribution of patients between center and noncenter (the sum of home hemodialysis and transplantation) therapies also differed significantly across the facilities (P < 0.001). Among white patients, therefore, the mode of therapy was dependent on the facility at which the patient received treatment. At the beginning of the study, 150 nonwhite patients were receiving maintenance therapy for ESRD (Table 2). Neither the ages of the patients nor their distribution by sex, travel time to treatment, marital status, work or employment status, or the presence of diabetes mellitus differed significantly across the four facilities (P > 0.05). Among the nonwhite patients, 107 (71. 3 %) were receivng center hemodialysis, 9 (6.0%) were on home hemodialysis, and 34 (22.7%) had functioning kidney transplants. Neither the distribution of patients among the three therapeutic modes nor the distribution of patients between center and noncenter therapies differed significantly across the facilities (P > 0.05). Thus, while obvious differences in the frequency of nonwhite patients receiving various modes of therapy were observed among the facilities, they were not statistically significant. In addition to the analysis of patient distribution among modes of maintenance therapy at the beginning of the study, the transfer of patients to alternate therapeutic modes during the observation peTable 3.
riod was also examined. Of the 202 patients receiving center hemodialysis at the initiation of the study, only 28 (13.9%) transferred to an alternate therapeutic mode (Table 3); the preponderance of those who transferred received a renal transplant (82.1 %). Furthermore, the mode of treatment to which center hemodialysis patients transferred differed significantly across the four facilities (P < 0.001). Thus, for example, the frequency of transfers to transplantation among center hemodialysis patients at facility C was 3.5 times greater than at facility A. Similarly, the frequency of transfers to home hemodialysis at facility B was 7.3 times greater than the frequency at facility A and 3.2 times greater than the mean for the four facilities. Since the ages of the center hemodialysis patients and their distribution by race, sex, travel time to treatment, marital status, work or employment status, and the presence of diabetes mellitus did not differ significantly among the facilities (P > 0.05), the type of transfer among center hemodialysis patients was dependent on the facility at which they were treated. Among the 55 patients receiving home hemodialysis at the beginning of the study, 11 (20.0%) transferred to an alternate therapy (Table 3). Similar to center hemodialysis patients, the majority of home patients who transferred received a renal transplant (72.7%). Of the 162 transplant recipients at the initiation of the study, 11 (6.8 %) rejected the graft during the observation period and required a return to hemodialysis. Among the 11 patients, however, only two did not pursue further renal transplantation. The insufficient size of these samples precluded a specific analysis of the potential facility dependence of transfers from home hemodialysis and transplantation. DISCUSSION
In their examination of dialysis and transplant rates, ReIman and Rennie 8 observed a substantial
ESRD Patients Transferring to Alternate Treatment Modes by Facility and Therapeutic Mode' CH Patients Transferring to
T Patients Transferring to
HH Patients Transferring to
Facility
HH
T
Total
CH
T
Total
CH
HH
Total
A
1.1 8.0 5.0 1.4 2.5
5.7 12.0 20.0 15.7 11.4
6.9 20.0 25.0 17.1 13.9
3.8 5.9 0 11.1 5.5
23.1 11.8 0 0 14.5
26.9 17.6 0 11.1 20.0
0 0 0 4.2 0.6
0 7.1 0 0 0.6
0 7.1 0 4.2 1.2
B C D Total
CH = center hemodialysis; HH * All numbers are percentages.
=
home hemodialysis; T
= transplantation.
25
TREATMENT BIAS IN ESRD
variation within both the United States and the international community. Similarly, data developed in the present study demonstrated a significant difference in the distribution of patients across ESRD treatment modes among individual facilities within the same health service area. While patients at .the four participating facilities were comparable on age, sex, travel time to treatment, marital status, work or employment status , and the presence of diabetes mellitus , a statistically significant difference in the racial composition of the facilities was also apparent. A more detailed analysis of the data revealed the mode of maintenance therapy to be facility-dependent among the white ESRD patients who comprised 64 .2 % of the patients studied. For example, among these patients the percentage receiving center hemodialysis at facility D was 2.6 times greater than at facility A (Table 2); the percentage of home hemodialysis patients was 3.2 times greater at facility B than at facility C. While a statistically significant difference in the frequency of nonwhite patients receiving various modes of ESRD therapy was not present across the facilities, substantial variation was nonetheless observed; for example, 22.2 % of the nonwhite patients at facility B were on home hemodialysis compared with 0 % at facility C. Although sociodemographic and diagnostic differences among the white patients might be proposed as an explanation,30 the data clearly did not support this argument. White patients treated at the four facilities were comparable on age, sex, travel time to treatment, marital status, work or employment status, and the presence of diabetes mellitus. This was not surprising since the facilities are located within 2.5 km of each other. It might also be hypothesized that home hemodialysis and renal transplantation were perceived as competitive therapies; thus , the selection of either therapeutic mode would merely reflect the physician or patient preference for a therapy of equivalent benefit. However, the statistically significant difference in the distribution of white patients between center and noncenter therapies across the facilities precluded this explanation. It would appear, therefore , that the bias observed in the treatment of these patients reflected the attitudes of the staff and patients toward the three modes of therapy at the respective facilities , rather than inherent sociodemographic or diagnostic differences in the patient populations served.
While the retrospective design of the study did not permit the direct examination of the importance of these attitudes, additional support for this conclusion was evidenced by the facility dependence of patient transfers from center hemodialysis to alternate modes of maintenance therapy. It is also noteworthy that the available evidence did not suggest that the treatment differences observed in this study were a consequence of the attitudes of referring physicians or their knowledge of the therapeutic preferences of individual facilities. Irrespective of the distribution of patients among treatment modes and the substantial progress evidenced by several facilities in achieving comparatively high rates of transfer to home hemodialysis and transplantation, it was equally apparent that patients continued their current mode of therapy. The stability of patients within each of the therapies suggested that adaptation 31 ' 36 was associated with specific therapeutic modes , rather than with ESRD therapy in general. The implications of these findings for both the formulation of public policy and the management of patient care are substantial. The probable relationship between staff attitudes and the selection of treatment modes suggests that further attention be directed toward improved professional education as well as peer support and review. Furthermore, the intensity of adaptation to specific modes of maintenance therapy implies that if home hemodialysis and transplantation are to be substantive components of a national ESRD policy, they should be employed, with increased frequency, as the initial modes of treatment. ACKNOWLEDGMENT The authors wish to express their appreciation to David J. Baltzer and Thomas J. Valerius for their invaluable assistance in the collection of the data presented in this study.
REFERENCES I. Hoffstein PA, Krueger KK , Wineman RJ: Dialysis costs: Results of a diverse sample study. Kidney Int 9:286-293, 1976 2. Burton BT: End-stage renal disease: Better treatment at lower cost. Ann Intern Med 88:567-568 , 1978 3. Adams LR: Medicare coverage for chronic renal disease: Policy implications. Health Soc Work 3:41-53, 1978 4 . Evans RW, Blagg CR , Bryan FA: Implications for health care policy: A social and demographic profile of hemodialysis patients in the United States. JAMA 245 :487-491 , 1981 5. Guttmann RD: Renal transplantation. N Engl J Med 301 :975-982 , 1979
26 6. Rettig RA: The politics of health cost containment: Endstage renal disease. Bull NY Acad Med 56:115-137, 1980 7. Stange PV, Sumner AT: Predicting treatment costs and life expectancy for end-stage renal disease. N Engl J Med 298:372-378, 1978 8. Reiman AS, Rennie D: Treatment of end-stage renal disease: Free but not equal. N Engl J Med 303:996-998, 1980 9. Roberts SD, Maxwell DR, Gross TL: Cost-effective care of end-stage renal disease: A billion dollar question. Ann Intern Med 92:243-248, 1980 10. Wineman RJ: End-stage renal disease: 1978. Dial Transplant 7:1034-1065,1978 11.Drees A, Gallagher EB: Hemodialysis, rehabilitation, and psychological support, in Levy NB (ed): Psychological Factors in Hemodialysis and Transplantation. New York, Plenum, 1981, pp 133-145 12. Blagg CR, Hickman RO, Eschbach JW, et al: Home hemodialysis: Six years' experience. N Engl J Med 283: 11261131, 1970 13. Cadnapaphornchai P, Kuruvila KC, Holmes J, et al: Analysis of five years' experience of home dialysis as a treatment modality for patients with end-stage renal failure. Am J Med 57:789-799, 1974 14. Roberts JL: Analysis and outcome of 1,068 patients trained for home hemodialysis. Kidney Int 9:363-374, 1976 15. Salvatierra 0: Renal transplantation in perspective. Dial Transplant 7:171-180,1978 16. Havry P: Dialysis and kidney transplantation in Europe. Med Bioi 55:245-248, 1977 17. Whalen JE, Freeman RM: Home hemodialysis review in Iowa. Arch Intern Med 138:1787-1790, 1978 18. Brenner BM, Lazarus JM: Chronic renal failure, in Isselbacher KJ, Adams RD, Braunwald E, et a1 (eds): Harrison's Principles ofInternal Medicine. St Louis, McGraw-Hill, 1980, pp 1299-1307 19. Blagg CR, Scribner BH: Dialysis: Medical, psychosocial, and economic problems unique to the dialysis patient, in Brenner BM, Rector FC (eds): The Kidney. Philadelphia, Saunders, 1976, pp 1705-1744 20. Robinson BHB: Intermittent haemodialysis in the home. Br Med Bull 27: 173-180, 1971
SMITH ET AL 21. Mock LA, Kopel K: Psychosocial aspects of home and in-center dialysis. Dial Transplant 6:40-43, 1977 22. Levine C: Dialysis or transplant: Values and choices. Hastings Cent Rep 8:8-10, 1978 23. Dutz H, Mebel M, Roseler E, et al: Evaluation of factors involved in rehabilitation using dialysis and transplantation. Proc 6th Int Congr Nephrol. Basel, Switzerland, Karger, 1976, pp 710-714 24. Anonymous: Selection of patients for dialysis and transplantation. Br Med J 2:1449-1450, 1978 25. Guttmann RD: Renal transplantation. N Engl J Med 301:1038-1048,1979 26. Levy NB, Abram HS, Kemph Jp, et al: Panel: Living or dying: Adaptation to hemodialysis, in Levy NB (ed): Living or Dying: Adaptation to Hemodialysis. Springfield, Ill, Charles C Thomas, 1974, pp 3-29 27. Greater St Louis Health Systems Plan: 1980-1984. St Louis, Greater St LOUIS Health Systems Agency, 1980, pp ESRD 1-13 28. Gutman RA, Stead WW, Robinson RR: Physical activity and employment status of patients on maintenance dialysis. N Engl J Med 304:309-313, 1981 29. Nie NH, Hull CH , Jenkins JG, et al: SPSS: Statistical Package for the Social Sciences (ed 2). New York, McGrawHill,1975 30. Alexandre GPJ: Factors involved in choosing dialysis or transplantation. Proc 6th Int Congr Nephrol, Basel, Switzerland, Karger, 1976, pp 692-697 31. Seime RJ, Zimmerman J: Adjustment to dialysis: A brief review. Behav Med Update 3:9-11, 1981 32. De-Nour AK, Shaltiel J, Czaczkes JW: Emotional reactions of patients on chronic hemodialysis. Psychosom Med 30:521-533, 1968 33. Levy NB: The psychology and care of the maintenance hemodialysis patient. Heart Lung 2:400-405, 1973 34. Flannery JG: Adaptation to chronic renal failure. Psychosomatics 19:784-787, 1978 35. Abram HS: The psychiatrist, the treatment of chronic renal failure, and the prolongation of life: II. Am J Psychiatry 126:157-166, 1969 36. Levy NB: Psychological sequelae to hemodialysis. Psychosomatics 19:329-331, 1978