HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Le Quintrec, J.-L. Articles by Piette, F. Search for Related Content PubMed PubMed Citation Articles by Le Quintrec, J.-L. Articles by Piette, F.

Randomized Controlled Drug Trials on Very Elderly Subjects: Descriptive and Methodological Analysis of Trials Published Between 1990 and 2002 and Comparison With Trials on Adults

¹ Service de Gérontologie I, Hôpital Sainte Périne/Rossini, Paris, France.
² Service de Gérontologie 3, Hôpital C. Foix, Ivry sur Seine, France.
³ Département de Biostatistiques, Faculté Pitié-Salpetrière, Paris, France.
⁴ Laboratoire d'Ethique Médicale, Université Paris 5, France.
⁵ Service de Médecine Interne, Hôpital C. Foix, Ivry sur Seine, France.

Address correspondence to Jean-Laurent Le Quintrec, Hôpital Sainte-Périne/Rossini, Service de Gérontologie I, 11, rue Chardon Lagache, F 75016 Paris, France. E-mail: muriel.palomares{at}spr.ap-hop-paris.fr

	Abstract

Top Abstract Methods Results Discussion References

 
Background. Very elderly subjects (VES; aged 80 years or older) constitute a special population as they frequently present multiple diseases (polypathology). Results from trials on general adult populations therefore cannot be extrapolated to VES. We performed a census of randomized controlled trials (RCT) on VES published between 1990 and 2002, and carried out a descriptive and methodological analysis of these RCT/VES, comparing them with matched RCT on general adult populations (control RCT, RCT/C).

Methods. We searched for RCT/VES in two international databases (EMBASE and MEDLINE) and then manually. RCT/C were matched to RCT/VES for disease area and year of publication. The methodological quality of each RCT was assessed with Chalmers' scale.

Results. We identified 84 RCT/VES, 63 of which were conclusive and 21, inconclusive. Subjects were institutionalized in 48 RCT, and community dwelling in 11 RCT (unspecified in 25 RCT). Efficacy was the main criterion in 75 RCT; tolerance in 9 RCT. Twenty-six RCT were published by geriatrics journals, and 58 by general medical journals. The RCT/VES covered most of the disease areas of geriatrics. The 84 RCT/VES had a mean methodological quality score of 0.578 ± 0.157. The matched 84 RCT/C had a mean methodological quality score of 0.592 ± 0.116 (p =.466). The methodological quality score of RCT/VES increased with the number of included subjects (p =.004) and the year of publication (p =.001).

Conclusions. The methodological quality of RCT/VES is equivalent to that of RCT in general adult populations. Nevertheless, RCT/VES remain very scarce, and neglect certain diseases. RCT/VES and the inclusion of very elderly subjects in RCT on adults should be strongly encouraged.

	METHODS

Top Abstract Methods Results Discussion References

 
We performed a census of RCT on drug therapy in subjects with a mean age over 80 years, published between January 1990 and December 2002. Our search was not limited to the English language. We began by searching two electronic databases: EMBASE and MEDLINE. We selected a large group of key words (randomized controlled trials, randomized controlled trials, double blind method, prospective studies, comparative studies, crossover studies, controlled clinical trials), which were used as "medical subject headings" and/or "text words." We then manually checked the references of each selected article.

We assessed the following characteristics of the RCT on VES (RCT/VES): aim of the trial (efficacy or tolerance), type of endpoint (major clinical or surrogate), result (positive or inconclusive), age and autonomy of included subjects, disease areas, type of journal publishing the RCT (geriatrics journal or general medical journal), number of included subjects, and methodological quality (as assessed with Chalmers' scale). A major clinical endpoint was defined as the "true endpoint" (13) for which the tested drug was developed, e.g., decrease in frequency of infections for prophylactic anti-infection drugs, decrease in frequency of fractures for antiosteoporotic drugs. Conversely, a surrogate endpoint was defined as a nonclinical and/or nonmajor endpoint, e.g., increase in biological immune markers for prophylaxis of infections, or increase in bone mineral density for antiosteoporotic drugs.

We then searched for control RCT (RCT/C) on subjects with a mean age <80 years. Each RCT/C was randomly selected from all RCT on adults (18 years or over) published in the same disease area and during the same year as the RCT/VES to which they were matched. For each RCT/C, we assessed aim of the trial, type of endpoint, results, mean age, number of included subjects, and methodological quality (using Chalmers' scale).

We assessed the methodological quality of RCT using Chalmers' scale (14). This scale, which was not specifically designed for geriatric RCT, is very extensive (15). It comprises 27 items grouped into three sections: the first section (14 items) assesses the design of the trial and is scored out of 60; the second section (9 items) assesses statistical analysis and is scored out of 30; and the third section (4 items) assesses presentation of the results and is scored out of 10. The final score obtained is therefore between 0/100 (lowest quality) and 100/100 (highest quality) and has been scaled to give a number between 0 and 1. Each article was assessed by two blinded independent geriatricians. The final score was obtained by consensus, as suggested by Chalmers (14).

Statistical analyses were performed in two steps. In the first, each group of trials—RCT/VES and RCT/C—was studied separately, to assess the relationships between methodological quality score and aim of the trial, type of endpoint, result of the trial, type of journal, number of subjects, and year of publication. All the comparisons were performed using nonparametric tests (Wilcoxon tests) for qualitative variables and Spearman correlation coefficient tests for quantitative variables. In the second step, RCT/VES and RCT/C were compared by means of pair-wise Student's t tests for quantitative variables and McNemar tests for qualitative variables. The SAS 8.1 package (SAS Institute, Cary, NC) was used for all statistical calculations. We used a significance threshold of p =.05 for all tests.

	RESULTS

Top Abstract Methods Results Discussion References

 
We identified 84 RCT/VES on drug therapy published between January 1990 and December 2002. These 84 RCT/VES covered most of the disease areas of geriatrics (Table 1). The aim of the RCT concerned drug efficacy in 75 RCT, and drug tolerance (Table 2) in nine. The endpoint of the trial was major clinical in 65 RCT, and surrogate in 19 RCT. Among the 84 RCT/VES, 63 led to a conclusive result, and 21 were inconclusive.

Fifty-eight of the RCT were published by general medical journals, and 26 by geriatrics journals. The number of included subjects was between 7 and 5125 (mean ± SD: 264 ± 715) and did not differ significantly according to the aim of the trial (p =.259), type of endpoint (p =.556), result (p =.430), or type of journal (p =.630).

The mean methodological quality score of the 84 RCT/VES was 0.578 ± 0.157 and did not differ significantly according to aim of the trial (p =.558), type of endpoint (p =.101), result (p =.098), or type of journal (p =.654). The methodological quality score of RCT/VES increased with the number of included subjects (p =.004) and with the year of publication (p =.001).

In the 84 RCT/C, the aim concerned drug efficacy in 73 RCT and drug tolerance in 11. The endpoint was major clinical in 62 RCT and surrogate in 22. Among the 84 RCT/C, 62 led to a conclusive result, and 22 were inconclusive. The mean age of included subjects was between 22.3 and 79.6 years (overall mean age of all subjects included: 57.2 ± 1.8 years). There was an upper age limit in 18 RCT/C.

The number of included subjects in the 84 RCT/C was between 10 and 21,106 (mean ± SD = 530 ± 2482). The number of included subjects was significantly higher in RCT/C with a major clinical endpoint than in those with a surrogate endpoint (p <.0001) but did not differ significantly according to the aim of the trial (p =.600), result (p =.202), or type of journal (p =.170).

The mean methodological quality score of the 84 RCT/C was 0.592 ± 0.116. The methodological quality score was significantly higher in RCT/C with a major clinical endpoint than in those with a surrogate endpoint (p <.0001), but did not differ significantly according to the aim of the trial (p =.883), result (p =.497), or type of journal (p =.210). The methodological quality score of the RCT/C increased with the number of included subjects (p =.0007) and with the year of publication (p =.006).

Comparisons between RCT/VES and RCT/C revealed no statistically significant difference for study protocol (Table 3). The number of included subjects was similar in RCT/VES and RCT/C (p =.757). Finally, the methodological quality was equivalent in RCT/VES and RCT/C (p =.466).

	DISCUSSION

Top Abstract Methods Results Discussion References

We searched for RCT/C using the same search strategy as for RCT/VES. RCT/C were matched with RCT/VES for year of publication, to avoid any discrepancy due to possible improvements in methodological quality between 1990 and 2002. RCT/C were also matched with RCT/VES for disease area, to allow valuable comparisons between RCT/VES and RCT/C. This strategy also eliminated bias due to journals, in that we searched for RCT/C not only in the journals that published RCT/VES, but in all journals dealing with the same disease area. The overall mean age of all subjects included in the 84 RCT/C was very different from that of all subjects in the 84 RCT/VES (57.2 ± 1.8 years vs 83.0 ± 2.0 years), and there was therefore a minimal age distribution overlap between the two groups of RCT.

Most of the RCT/VES were designed to determine drug efficacy, which is the primary goal for RCT. Tolerance is also a major concern in geriatrics, because the risk/benefit ratio of treatments in VES differs greatly from that in adults as a whole (10). This difference precludes the extrapolation of results obtained in populations of adults. In our study, only 9 RCT/VES and 11 RCT/C aimed to assess drug tolerance. The proportion of RCT/VES aimed at assessing drug tolerance would therefore seem to be insufficient, given that the prevalence of multiple chronic conditions and subsequent polymedication, generating numerous side effects, increases with age (18). There is also a need for RCT identifying the minimum effective dose of some drugs used in VES, especially drugs for Parkinson's disease, diabetes, hypertension, and coronary disease (19).

The endpoint of RCT should be clinical, as much as possible, because only major clinical endpoints can demonstrate the real clinical efficacy of the drug tested. In some disease areas (e.g., osteoporosis) and for preventive treatments (e.g., cardiovascular diseases and cancer), the clinical endpoint requires long-term follow-up and a large number of included subjects; neither of these conditions is easy to fulfill with VES. Surrogate endpoints may be used in such cases, provided that the surrogate endpoint has been shown to be linked to the corresponding clinical endpoint. Nineteen of the 84 RCT/VES were based on surrogate endpoints. Most of these 19 RCT/VES dealt with disease prevention—for osteoporosis (7 RCT), infections (7 RCT), nutritional problems (3 RCT), and serum lipid concentrations (1 RCT). The proportion of trials with surrogate endpoints was similar in RCT/VES and RCT/C. Given the difficulties involved in carrying out RCT/VES with a major clinical endpoint, geriatricians cannot be considered to "overuse" surrogate endpoints.

Twenty-one RCT/VES were inconclusive: this is an important finding because it was thought until recently that pharmaceutical companies and editorial boards avoided publishing inconclusive trials (20). The distribution of inconclusive RCT/VES between disease areas was homogeneous and did not differ from that of conclusive RCT/VES. The proportion of inconclusive trials among RCT/VES was similar to that of inconclusive trials among RCT/C. The publication of inconclusive trials therefore seems to be becoming more common for studies of both geriatric and general adult populations; this trend should be encouraged.

Despite the high mean age of included subjects, 11 RCT/VES were devoted to community-dwelling subjects. Most of these 11 RCT dealt with disease prevention (5 RCT) or preservation of autonomy (4 RCT), both of which can only effectively be studied in autonomous subjects living in the community. In 25 RCT/VES, the level of autonomy of the included subjects was not specified. This lack of detail is a particularly serious omission in RCT in geriatrics, because VES constitute a very heterogeneous population. In this population, age alone is insufficient for the assessment of autonomy and health because of differential aging: some people aged 80 or older live in the community like other adults, whereas others suffer multiple diseases and live in nursing homes. It is therefore essential for both scientific and ethical reasons to report the place of residence or lifestyle in RCT/VES, because these factors define the autonomy or polypathology of the VES included in the trial.

Two major disease areas are neglected by RCT/VES: Parkinson's disease and cancer. For Parkinson's disease, Mitchell and colleagues (2) conducted a systematic search of all RCT published between 1966 and 1996. Of the 112 articles they identified, 42 included subjects older than 75 years, and only eight adequately reported the age groups of the included subjects. In our search, we found no RCT on drugs to treat Parkinson's disease in subjects aged 80 or older published between 1990 and 2002. The exclusion of elderly subjects from cancer research has been criticized since the 1980s. In 1987, Begg and Carbone (5) showed that patients aged 70 years or older had rates of response and toxicity similar to those in younger adults, and agreed with the National Cancer Institute statement that age is not a criterion for the exclusion of patients. In 1994, Trimble and colleagues (6) found that "with respect to cancer incidence, older patients (65 years and older) are underrepresented in cancer treatment trials." Montfardini and colleagues (7), in phase II trials including 2344 patients, found that 22% of patients were aged 65 years or older and only 8% of patients were aged 70 years or older.

Fifty-eight of the 84 RCT/VES were published by general medical journals. Clinical trials on geriatric topics are no longer confined to geriatrics journals but are now considered interesting by the editorial boards of general medical journals. The RCT/VES published by geriatrics journals were mainly devoted to "specific" geriatric disease areas such as urinary incontinence, Alzheimer's disease, nutritional problems, and immune function.

The mean methodological quality score of the 84 RCT/VES was 0.578 ± 0.157. This score is very different from the perfect score of 1 for the Chalmers' scale. This low mean score may be due to the nature of the Chalmers' scale, which was not specifically designed for geriatric research. However, there are two main reasons why this is unlikely to be the case. First, although clinical research on VES certainly requires some adaptation of traditional methodologies and modification of expectations (12), RCT/VES are above all RCT, and therefore need to satisfy the same methodological quality criteria as RCT on any other subject. Second, the mean score of the RCT/C (0.592 ± 0.116) was very similar to that of RCT/VES (0.578 ± 0.157) (p =.466): the same search applied to RCT on VES and matched RCT on general adult populations gave similar results. The Chalmers' scale is therefore not biased in geriatric clinical trials.

This analysis of RCT/VES published between 1990 and 2002 highlights the rarity of RCT/VES. Clinical research on VES involves specific difficulties, which must be resolved if RCT are to be conducted in such populations. The principal difficulty is the great heterogeneity of VES, due to differential aging and multiple diseases. The recruitment strategy should therefore be broad, to avoid selecting subjects presenting only one disease; these subjects may not constitute a representative sample of the target population for the drug studied (18). So, inappropriate inclusion criteria may result in nonrepresentative samples of patients, as has been demonstrated in cancer research (9,21–23). There should therefore be very few exclusion criteria, and these should be based on ethical or pathophysiological grounds (1,9,10,23,24,25): inability to obtain informed consent from subject; imminent death of subject; contraindication of subject to the drug studied. Age itself should no longer be considered an exclusion criterion.

The second difficulty is that patients or their proxies are more reluctant to participate in RCT if the patient is very old and suffers from multiple diseases. This reluctance is often shared by the patient's personal physician (26). These difficulties may result in the recruitment of less disabled patients, who do not constitute a representative sample of the population. Such bias has been demonstrated in studies of elderly patients with acute myocardial infarction or lymphoma (21).

The third difficulty is the limited life expectancy of VES. The number of deaths should be estimated before enrollment is begun. This restricted life expectancy makes it necessary to choose an appropriate endpoint: morbidity and mortality endpoints are often not suitable for VES (2), especially in clinical trials dealing with preventive or symptomatic treatments. In these cases, indices of quality of life and activities of daily living are required (27).

Frequent cognitive and sensory deterioration in VES is also a difficulty, resulting in specific problems, such as poor compliance during the RCT (28), and may necessitate changes to the number and circumstances of clinical follow-up appointments (29). These difficulties are particularly great in clinical research focusing on Alzheimer's disease patients; this research should be restricted to drugs aimed at Alzheimer's disease treatment.

Our descriptive and methodological analysis showed a) that RCT/VES published since 1990 have a methodological level equivalent to that of RCT on general adult populations, and b) that methodological quality has increased over time. This result is very encouraging for geriatricians. Clinical research in geriatrics should now be aimed at performing RCT on VES, who constitute a growing population of patients suffering from cancer (30), Parkinson's disease (2), and cardiovascular disease (31), in particular.

Furthermore, researchers testing a new treatment on adults must include VES in RCT. As stated by the International Conference of Harmonization (ICH) Geriatric Guidelines (32), VES must be included in sufficient numbers in all RCT testing drugs "that are likely to have significant use in the elderly ... The older the population likely to use the drug, the more important it is to include the very old." Geriatricians, researchers, regulatory authorities, and pharmaceutical companies now have to translate these guidelines into facts.

	Footnotes

 
Decision Editor: John E. Morley, MB, BCh

Received July 17, 2003

Accepted October 3, 2003

	References

Top Abstract Methods Results Discussion References

 

1. Gurwitz JH, Col NF, Avorn J. The exclusion of the elderly and women from clinical trials in acute myocardial infarction. JAMA. 1992;268:1417-1422.[Abstract]
2. Mitchell SL, Sullivan EA, Lipsitz LA. Exclusion of elderly subjects from clinical trials for Parkinson's disease. Arch Intern Med. 1997;157:1393-1398.
3. Bugeja G, Kumar A, Banerjee AK. Exclusion of elderly people from clinical research: a descriptive study of published reports. BMJ. 1997;315:1059.[Free Full Text]
4. Bayer A, Tadd W. Unjustified exclusion of elderly people from studies submitted to research ethics committee for approval: descriptive study. BMJ. 2000;321:992-993.[Free Full Text]
5. Begg CB, Carbone PP. Clinical trials and drug toxicity in the elderly: the experience of the Eastern Cooperative Oncology Group. Cancer. 1983;52:1986-1992.[Medline]
6. Trimble EL, Carter CL, Cain D, et al. Representation of older patients in cancer treatment trials. Cancer. 1994;74:2208-2214.[Medline]
7. Montfardini S, Sorio R, Boes GH, et al. Entry and evaluation of elderly patients in European Organization for Research and Treatment of Cancer (EORTC) new drug-development studies. Cancer. 1995;76:333-338.[Medline]
8. Hutchins LF, Unger JM, Crowley JJ, et al. Underrepresentation of patients 65 years of age or older in cancer treatment trials. N Engl J Med. 1999;341:2061-2067.[Abstract/Free Full Text]
9. Stephens RJ, Johnson DH. Treatment and outcomes for elderly patients with small cell lung cancer. Drugs Aging. 2000;17:229-247.[Medline]
10. Gurwitz JH, Goldberg RJ, Gore JM. Coronary thrombolysis for the elderly? JAMA. 1991;265:1720-1723.[Medline]
11. Bell JA, May FE, Stewart RB. Clinical research in the elderly: ethical and methodological considerations. Drug Intell Clin Pharm. 1987;21:1002-1007.[Abstract]
12. Wilder Zimmer A, Calkins E, Hadley E, et al. Conducting clinical research in geriatric populations. Ann Intern Med. 1985;103:276-283.
13. Prentice RL. Surrogate endpoints in clinical trials: definition and operational criteria. Stat Med. 1989;8:431-440.[Medline]
14. Chalmers TC, Smith H, Jr, Blackburn B, et al. A method for assessing the quality of randomized control trial. Control Clin Trials. 1981;2:31-49.[Medline]
15. Moher D, Jadad AR, Graham N, et al. Assessing the quality of randomized controlled trials: an annotated bibliography of scales and checklists. Control Clin Trials. 1995;16:62-73.[Medline]
16. Nieminen P, Isohanni M. Bias against European journals in medical publication databases. Lancet. 1999;353:1592.[Medline]
17. Avenell A, Handoll H, Grant AM. Lessons for search strategies from a systematic review, in The Cochrane Library, of nutritional supplementation trials in patients after hip fracture. Am J Clin Nutr. 2001;73:505-510.[Abstract/Free Full Text]
18. Wolff JL, Starfield B, Anderson G. Prevalence, expenditures, and complications of multiple chronic conditions in the elderly. Arch Intern Med. 2002;162:2269-2276.[Abstract/Free Full Text]
19. Rochon PA, Tu JV, Anderson GM, et al. Rate of heart failure and 1-year survival for older people receiving low dose ß-blockers therapy after myocardial infarction. Lancet. 2000;356:639-644.[Medline]
20. Reidenberg MD. Decreasing publication bias. Clin Pharmacol Ther. 1998;63:1-3.[Medline]
21. Chen CI, Skingley P, Meyer RM. A comparison of elderly patients with aggressive histology lymphoma who were entered or not entered on to a randomized phase II trial. Leuk Lymphoma. 2000;38:327-334.[Medline]
22. Yancik R, Wesley MN, Ries LAG, et al. Effect of age and comorbidity in postmenopausal breast cancer patients aged 55 years and older. JAMA. 2001;285:885-892.[Abstract/Free Full Text]
23. Britton A, McKee M, Black N, et al. Threats to applicability of randomised trials: exclusions and selective participation. J Health Serv Res Policy. 1999;4:112-121.[Medline]
24. Lechat PH, Lardoux H, Mallet A, et al. Etude de l'association anticoagulant (fluindione)–aspirine, chez les patients en fibrillation auriculaire à haut risque de complications thrombo-emboliques. Une étude randomisée (FFAACS). Thérapie. 2000;55:681-689.[Medline]
25. Weijer C, Freedman B, Shapiro S, et al. Assessing the interpretation of criteria for clinical trial eligibility: a survey of oncology investigators. Clin Invest Med. 1998;21:17-26.[Medline]
26. Jeffe DB, Binder EF, Willinams DB, et al. Frail older women's participation in a trial of hormone replacement therapy: perceived benefits and concerns. Menopause. 2001;8:127-134.[Medline]
27. Wenger NK. Exclusion of the elderly and women from coronary trials: is their quality of care compromised? JAMA. 1992;268:1460-1461.[Medline]
28. Buist DSM, Lacroix AZ, Black DM, et al. Inclusion of older women in randomized clinical trials: factors associated with taking study medication in the Fracture Intervention Trial. J Am Geriatr Soc. 2000;48:1126-1131.[Medline]
29. Berkman CS, Leipzig RM, Greenberg SA, et al. Methodologic issues in conducting research on hospitalized older people. J Am Geriatr Soc. 2001;49:172-178.[Medline]
30. Kennedy BJ. Needed: clinical trials for older patients. J Clin Oncol. 1991;9:718-720.[Medline]
31. Gollub SB. Is intensive drug therapy appropriate for older patients? Lancet. 1999;353:940-941.[Medline]
32. The European Agency for the Evaluation of Medicinal Products. International Conference of Harmonization Topic E7. Note for guidance on studies in support of special population: Geriatrics (CPMP/ICH/379-95). London, UK: ICH Technical Coordination.

This article has been cited by other articles:

				Functional Outcomes for Clinical Trials in Frail Older Persons: Time To Be Moving: Working Group on Functional Outcome Measures for Clinical Trials J. Gerontol. A Biol. Sci. Med. Sci., February 1, 2008; 63(2): 160 - 164. [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Le Quintrec, J.-L. Articles by Piette, F. Search for Related Content PubMed PubMed Citation Articles by Le Quintrec, J.-L. Articles by Piette, F.