Quality of life in patients with lower limb ischemia; revised suggestions for analysis

Article Outline

• Abstract
• Methods
  • Short Form 36
  • Short Form 8
• Definitions
  • Validity
  • Reliability
  • Responsiveness
  • Statistical analysis
• Results
  • Validity (Table II)
  • Reliability (Table III)
  • Responsiveness (Fig 1, Table IV)
• Discussion
• Author contributions
• References
• Copyright

Background

Quality of life (QoL) is a crucial outcome measure in patients with lower limb ischemia (LLI). The Short Form 36 (SF36) has been proposed as the gold standard instrument for generic QoL analysis in patients with LLI. The Short Form 8 (SF8) was developed from the SF36 and we aim to compare these two instruments in terms of validity, reliability, and responsiveness.

Methods

One hundred ninety-three patients, 135 men and 58 women, median age 66 (range, 44-84) years with LLI completed the SF36 and the SF8. Disease severity was graded according to International Society of Cardiovascular Surgery (ISCVS) suggested reporting standards. Correlation between the two instruments' like domains and non-like domains reflects convergent and divergent validity respectively. A subgroup of 58 patients (44 men) completed two sets of questionnaires, with an intervening period of 2 weeks. Correlation between these two sets of questionnaires was used to analyze test/retest reliability. Spearman's rank correlation was used to analyze validity and reliability. Responsiveness of the individual domains across the whole group was analyzed using the Kruskall-Wallis analysis of variance (ANOVA) test, while responsiveness between the groups of patients with varying severity of LLI was analyzed using the Mann-Whitney U test.

Results

There was greater correlation between like domains of SF36 and the SF8 than the non-like domains suggesting good convergent–divergent validity. Test/retest reliability was significant for both instruments (r_s > 0.7). Increasing LLI resulted in a statistically significant deterioration in all eight domains of both instruments. The time taken was significantly shorter and less assistance was required to complete the SF8 than the SF36.

Conclusion

The SF8 is a valid and reliable QoL instrument in patients with LLI, and as it is simpler and quicker to complete, we suggest it may challenge the SF36 as the gold standard generic QoL analysis instrument in LLI.

Quality of life (QoL) analysis is an essential outcome measure in chronic disease.¹ During the last decade, numerous studies have investigated the role of QoL analysis in lower limb ischemia (LLI) and it is now recognized as the most important primary endpoint in evaluating the effect of treatment in patients with LLI.², ³, ⁴, ⁵, ⁶, ⁷, ⁸, ⁹ Quality of life may be analyzed using either generic or disease-specific instruments.¹⁰, ¹¹ Generic instruments typically analyze global QoL domains including physical, social, and psychological well being. The main advantage of generic QoL instruments is their “off the shelf” availability, allowing comparison between differing disease states and populations. The medical outcome study (MOS) 36-item short form (SF36) health survey is the most widely used generic QoL tool and it has been rigorously tested for validity, reliability, and responsiveness.¹², ¹³, ¹⁴, ¹⁵ Ten years ago, the SF36 was recommended as the gold standard European generic QoL instrument against which other instruments may be gauged.¹⁶ However, the SF36 is lengthy, time consuming to complete and score, and associated with a significant incidence of incomplete responses. These problems represent practical obstacles to its routine use in clinical practice. In an attempt to address these problems, a new instrument, the short form 8 (SF8) has been developed from the SF36, with a reduction in items from 36 to 8.¹² We aimed to assess whether the SF8 was as valid, reliable, and responsive as the SF36 in patients with LLI and could, therefore, potentially replace the SF36 in this patient group.

Back to Article Outline

Methods 

Consecutive patients presenting to the Academic Vascular Surgical Unit at Hull Royal Infirmary between August and December 2006 with varying degrees of LLI were invited to participate in the study. In total, 193 patients were studied (135 men, 58 women), median age 66 (range, 44-84) years. Patients with intermittent claudication completed a treadmill test (speed 2.5 km/hour; gradient 10°). Pre and post treadmill ankle brachial pressure indices (ABPI) were recorded and the patient's disease severity graded according to the International Society of Cardiovascular Surgery (ISCVS) suggested reporting standard (30 mild, 52 moderate, 73 severe claudicants, 16 rest pain, and 21 tissue loss – Table I).¹⁷ All patients completed both the SF36 and SF8 at the time of the assessment and the time taken to complete each questionnaire was noted. The number of questionnaires either incompletely or erroneously answered was noted and the patient was then assisted to complete the questionnaire fully.

Table I. Patient demographics and clinical indicators of LLI

	Claudication
	Mild	Moderate	Severe	Rest pain	Tissue loss	P value
n	30	52	73	16	21
Median age (IQR)/years	64(56-73.5)	67.5(64-71)	65(60-74)	70(66-73)	75.5(69.5-79)	<.001
Male:female ratio	23:7	39:13	51:22	10:6	12:10	<.001
Median ABPI (IQR)	0.95(0.83-1.0)	0.93(0.69-1.0)	0.65(0.49-0.8)	0.38(0.23-0.46)	0.22(0.18-0.42)	<.001

IQR, Inter quartile range; ABPI, ankle brachial pressure index; n, number; LLI, lower limb ischemia. Statistical analysis undertaken using Kruskall-Wallis analysis of variance (ANOVA).

Short Form 36 

The SF36 was designed to provide assessment of generic health domains that are not specific to age, disease, or treatment group.¹³ It contains 36 items covering eight health-related QoL domains: bodily pain (BP), physical functioning (PF), role limitations due to physical problems (RP), role limitations due to emotional problems (RE), vitality (V), social function (SF), mental health (MH), and general health (GH). For each domain, questions are scored, coded, summed, and transformed on a scale from 0 (worst health) to 100 (best health). In this study, the United Kingdom version, two of the SF36s were used. The SF36 has shown acceptable validity and reliability in population studies and in patients with LLI.¹⁵, ¹⁶

Short Form 8 

The SF8 was constructed to replace the SF36 in population health surveys. It has been translated and linguistically validated for use in more than 30 languages. The SF8 is an eight-item version of the SF36 that yields a comparable eight-dimension health profile. Each SF8 single-item scale can be scored on the same norm-based metrics as the SF36 scales. The SF8 has demonstrated acceptable validity and reliability in population studies.¹²

Back to Article Outline

Definitions 

Validity 

Validity is defined as “the extent to which an instrument measures what it purports to measure”.¹⁶ Convergent validity assesses the correlation between like domains, while divergent validity assesses the correlation between non-like domains of the SF36 and SF8 QoL instruments. Spearman's rank correlation has been used to assess correlation between domains. Using the analysis method outlined above, the demonstration of validity requires the correlation between “like” domains measured by the different instruments to be greater than the correlation between non-like domains, ie, the correlation between SF8-measured pain and SF36-measured pain should be greater than the correlation between SF8-measured pain and the other SF36-measured domains.

Reliability 

Reliability is defined as the “extent to which an instrument yields the same results on the same population under different conditions”.¹⁶ Test/retest reliability is reflected in the correlation between scores at two time periods. Fifty-eight patients (44 men) completed two sets of questionnaires. The first set was completed at the time of the initial assessment and the second set completed 2 weeks later and returned by post. No interventions were undertaken between these two time points. Forty-two patients with claudication and 35 patients with critical limb ischemia were sent repeat questionnaires. All the claudicants and 16 critical limb ischemic patients returned completed forms.

Responsiveness 

Responsiveness is the ability to detect small but clinically significant changes.¹⁶ Responsiveness was determined by observing changes in the domain scores of both QoL questionnaires as lower limb ischemia became more severe.

Statistical analysis 

Data analysis was performed using SPSS v.13.0.0 for Windows (SPSS Inc, Chicago, Ill). Convergent/divergent validity and test/retest reliability of the QoL instruments was analyzed using the Spearman's rank correlation test. Kruskall-Wallis analysis of variance (ANOVA) was used to analyze changes in individual QoL domains across the whole group, while the Mann-Whitney U test was used to analyze responsiveness between the groups of patients with varying severity of LLI. A P value of < .05 was taken to represent statistical significance.

Back to Article Outline

Results 

Increasing LLI is associated with a reduction in the male to female ratio of patients affected, a significant increase in patients' age and a significant fall in ABPI (Table I). Increasing LLI resulted in a significant deterioration in all eight domains of both the SF8 and the SF36 (P < .05, Kruskall-Wallis ANOVA – Fig 1). The time taken to complete the SF8 (median 2.5 minutes, range, 1.5-3.5 minutes) was significantly shorter than the time taken to complete the SF36 (median 11 minutes, range, 8-20 minutes; P < .001 Wilcoxon Rank test). Twenty-seven percent of patients required assistance to complete the SF36 questionnaire as they found some questions confusing, whereas none of the patients required assistance in completing the SF8 questionnaire (P = .001 χ² test).

• 
  • View Large Image
  • Download to PowerPoint

• Fig 1. 

  a, Variation in the SF36 quality of life domains with increasing lower limb ischaemia. b, Variation in theSF8 quality of life domains with increasing lower limb ischaemia.

Validity (Table II) 

Convergent validity was greater than divergent validity for six SF36 domains (PF, RP, GH, V, SF, MH) and for seven SF8 domains (PF, RP, BP, GH, V, SF, MH).

Table II. Convergent (bold) and divergent validity: correlation coefficients for domains measured by the SF36 and the SF8

PF, Physical functioning; BP, bodily pain; RP, physical problems; GH, general health; V, vitality; SF, social function; RE, emotional problems; MH, mental health; SF8, short form 8; SF36, short form 36.

Reliability (Table III) 

Test/retest reliability was excellent for both the SF36 (0.76< r_s <0.86) and the SF8 (0.8< r_s <0.88) with the SF8 in general showing higher correlation coefficients.

Table III. Test/retest reliability coefficients (r_s) for the SF36 and the SF8

(r_s) – Spearman's rank correlation coefficient.

Responsiveness (Fig 1, Table IV) 

For the SF36, significant differences were seen between mild claudicants and all other groups of patients with LLI. No significant difference between moderate and severe claudicants was seen for any domain. No significant difference was seen between moderate claudicants and those patients with critical limb ischemia for the domains of GH and V, although there were significant differences seen between severe claudicants and patients with critical limb ischemia for these two domains. No differences were seen in domain scores when comparing patients with rest pain and patients with tissue loss.

Table IV. Responsiveness analysis for SF36 and SF8 instruments for LLI

	1v2	1v3	1v4	1v5	2v3	2v4	2v5	3v4	3v5	4v5	Total
SF36 Domains
Physical Functioning	X	X	X	X	NS	X	X	X	X	NS	8/10
Role Physical	X	X	X	X	NS	X	X	X	X	NS	8/10
Bodily Pain	X	X	X	X	NS	X	X	X	X	NS	8/10
General Health	X	NS	X	X	NS	NS	NS	X	X	NS	5/10
Vitality	X	X	X	X	NS	NS	NS	X	X	NS	6/10
Social Functioning	X	X	X	X	NS	X	X	X	X	NS	8/10
Role Emotional	NS	X	X	X	NS	X	X	NS	NS	NS	5/10
Mental Health	X	X	X	X	NS	X	X	X	X	NS	8/10
SF8 Domains
Physical Functioning	X	X	X	X	NS	X	X	X	X	NS	8/10
Role Physical	X	X	X	X	NS	X	X	X	X	NS	8/10
Bodily Pain	X	X	X	X	NS	X	X	X	X	NS	8/10
General Health	X	X	X	X	NS	NS	NS	NS	NS	NS	4/10
Vitality	X	X	X	X	NS	NS	NS	NS	NS	NS	4/10
Social Functioning	NS	NS	X	X	NS	X	X	X	X	NS	6/10
Role Emotional	X	X	X	X	NS	X	X	X	X	NS	8/10
Mental Health	X	X	X	X	NS	NS	NS	X	NS	NS	5/10

LLI, Lower limb ischemia; X: P < .05, (Mann-Whitney U test). NS, not significant, Total: number of significant differences between groups. ISCVS disease classification: 1 = mild, 2 = moderate, 3 = severe, 4 = rest pain, 5 = tissue loss.

For the SF8, mild claudicants had significant differences with all the other groups of patients except for the domain of social function. No significant differences were seen between moderate and severe claudicants for any domains. In the SF8 domains of GH, V, and MH, no significant differences were seen between patients with moderate/severe claudicants and those with critical limb ischemia. Furthermore, no differences were seen in domain scores when comparing patients with rest pain and patients with tissue loss. In summary, the SF8 is as responsive as the SF36 in three of the eight QoL domains (PF, RP, and BP) and more responsive in one QoL domain (RE) but less responsive in four QoL domains (GH, V, MH, and SF).

Back to Article Outline

Discussion 

QoL analysis is now an integral outcome measure in LLI.⁶, ⁹ Initial studies concentrated predominantly upon generic QoL tools with most using the SF36 as their preferred instrument.³, ¹⁶ The SF36 is without doubt the most widely used generic QoL instrument worldwide allowing comparison between different disease states and populations. The SF36 has also been recommended as the “gold standard” generic QoL instrument in the assessment of patients with LLI.¹⁶ However, critics of the SF36 claim it is prolonged, takes excessive time to complete, and generates an unacceptably high incidence of incomplete data sets.¹⁸ The ideal QoL instrument, in addition to being valid and reliable, should be sufficiently brief to achieve high response and completion rates, although sufficiently detailed to ensure adequate responsiveness. It should be simple for patients to comprehend and complete, and for health professionals to administer, score, and analyze. On this basis, the SF8 has been developed as a more concise version of the SF36 allowing both brevity and comprehensiveness in population health surveys.¹² By having only eight items, the SF8, therefore, relies on a single item to measure each of the eight domains in the SF-36. The recent development of the SF8 means that the instrument currently lacks the depth of validation data that is available for the SF36, data that is necessary before it can be recommended for routine practice.

Our results show that convergent/divergent validity is good for both the instruments but seems stronger (“like-domains” correlation coefficient greater than “unlike”) for the SF8 (seven of eight domains) than the SF36 (six of the eight domains). Reliability is strong for both QoL instruments but the correlation coefficients were on the whole higher for the SF8 than the SF36. The SF8 is a much shorter questionnaire, and as such, there may be an element of patient recall rather than reliability per se. However, previous work has shown that in a similar cohort of patients, the use of three differing lengths of QoL questionnaires resulted in similar test/retest reliability results suggesting that the results we observed were likely a true reflection of reliability.¹⁶

This study also confirmed previous findings that increasing LLI has a progressively detrimental impact on patient reported QoL, but this is the first study to demonstrate these findings using the SF8.¹⁶ No difference was observed in any of the SF36 or the SF8 domain scores between moderate and severe claudicants or patients with rest pain and tissue loss. This would suggest that when treating claudicants, an improvement to, at minimum, mild claudication is required and patients with critical limb ischemia need not only tissue healing but also absolution of rest pain to register QoL improvement. In comparison to the SF36, the SF8 demonstrated slightly inferior responsiveness in four domains, namely GH, SF, V, and MH. However, as these domains are frequently unaffected by small shifts in LLI severity, the importance of this finding is contentious and may be compensated for by the SF8's simplicity, brevity, patient acceptability, and completion rates. Of specific interest is the difference between domain scores in moderate claudicants and those patients with critical limb ischemia. The only difference between the two QoL tools is the lack of difference in the domain MH. However, an alternative health domain to measure psychological status is the RE domain that does show a significant difference between these two groups of patients.

The importance of QoL analysis in LLI is also evident by the number of disease-specific QoL tools that have been developed over the last decade and, on the whole, disease-specific QoL instruments may be more responsive than generic QoL instruments as well as having greater validity.¹¹, ¹⁹ It has been suggested that the King's College Hospital's Vascular Quality of Life Questionnaire (VASCU-QoL) is the only disease-specific QoL instrument that merits routine use in claudicants.¹⁹ However, generic QoL analysis is still essential in LLI, to enable comparisons between differing chronic disease states and as a basis for economic evaluation of intervention. We ourselves have not analyzed the effect of intervention in this cohort of patients. Previous work undertaken using the SF36 has shown that this tool is responsive to the effects of successful intervention in patients with LLI.²⁰, ²¹ Interventional claudication studies are ongoing using the SF8, but given the data presented in this paper, the SF8 may be expected to be as responsive as the SF36 in this regard.

Our results show that the SF8 is, on the whole, as valid, reliable, and responsive as the SF36, but importantly takes less time to complete, appears to be less confusing, and is, therefore, likely to be more acceptable to the patient across the disease spectrum of LLI. We therefore suggest that SF8 may challenge the SF36 as the gold standard generic QoL analysis instrument in LLI.

Back to Article Outline

Author contributions 

Back to Article Outline

References 

1. Idler EL, Benyamini Y. Self-rated health and mortality: a review of 27 community studies. J Health Soc Behav. 1997;38:21–37
   • View In Article
   • MEDLINE
   • CrossRef
2. Currie IC, Wilson YG, Baird RN, Lamont PM. Treatment of intermittent claudication: the impact on quality of life. Eur J Vasc Endovasc Surg. 1995;10:356–361
   • View In Article
   • Abstract
   • Full-Text PDF (491 KB)
   • CrossRef
3. Pell JP. Impact of intermittent claudication on quality of life (The Scottish Vascular Audit Group). Eur J Vasc Endovasc Surg. 1995;9:469–472
   • View In Article
   • Abstract
   • Full-Text PDF (344 KB)
   • CrossRef
4. Klevsgard R, Hallberg IR, Risberg B, Thomsen MB. Quality of life associated with varying degrees of chronic lower limb ischaemia: comparison with a healthy sample. Eur J Vasc Endovasc Surg. 1999;17:319–325
   • View In Article
   • Abstract
   • Full-Text PDF (171 KB)
   • CrossRef
5. Khaira HS, Hanger R, Shearman CP. Quality of life in patients with intermittent claudication. Eur J Vasc Endovasc Surg. 1996;11:65–69
   • View In Article
   • Abstract
   • Full-Text PDF (451 KB)
   • CrossRef
6. Schroeder TV. The TASC supplement - international recommendations for management of peripheral arterial disease. Eur J Vasc Endovasc Surg. 2000;19:563
   • View In Article
   • Full-Text PDF (34 KB)
   • CrossRef
7. Cassar K, Bachoo P, Brittenden J. The effect of peripheral percutaneous transluminal angioplasty on quality of life in patients with intermittent claudication. Eur J Vasc Endovasc Surg. 2003;26:130–136
   • View In Article
   • Abstract
   • Full Text
   • Full-Text PDF (123 KB)
   • CrossRef
8. Chetter IC, Spark JI, Kent PJ, Berridge DC, Scott DJ, Kester RC. Percutaneous transluminal angioplasty for intermittent claudication: evidence on which to base the medicine. Eur J Vasc Endovasc Surg. 1998;16:477–484
   • View In Article
   • Abstract
   • Full-Text PDF (1154 KB)
   • CrossRef
9. Dormandy JA, Rutherford RB. Management of peripheral arterial disease (PAD) (TASC Working Group. TransAtlantic Inter-Society Consensus (TASC)). J Vasc Surg. 2000;31(1 Pt 2):S1–S296
   • View In Article
   • MEDLINE
   • CrossRef
10. Chetter IC, Scott DJ, Kester RC. An introduction to quality of life analysis: the new outcome measure in vascular surgery. Eur J Vasc Endovasc Surg. 1998;15:4–6
    • View In Article
    • Full-Text PDF (318 KB)
    • CrossRef
11. Mehta T, Venkata Subramaniam A, Chetter I, McCollum P. Disease-specific quality of life assessment in intermittent claudication: review. Eur J Vasc Endovasc Surg. 2003;25:202–208
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (112 KB)
    • CrossRef
12. The SF community: offering information and discussion on health outcomes. http://www.sf-36.org/
    • View In Article
13. Ware JE, Sherbourne CD. The MOS 36-item short-form health survey (SF-36) (I. Conceptual framework and item selection). Med Care. 1992;30:473–483
    • View In Article
    • MEDLINE
    • CrossRef
14. Stewart AL, Hays RD, Ware JE. The MOS short-form general health survey (Reliability and validity in a patient population). Med Care. 1988;26:724–735
    • View In Article
    • MEDLINE
    • CrossRef
15. Brazier JE, Harper R, Jones NM, O'Cathain A, Thomas KJ, Usherwood T, et al. Validating the SF-36 health survey questionnaire: new outcome measure for primary care. BMJ. 1992;305:160–164
    • View In Article
    • MEDLINE
    • CrossRef
16. Chetter IC, Spark JI, Dolan P, Scott DJ, Kester RC. Quality of life analysis in patients with lower limb ischaemia: suggestions for European standardisation. Eur J Vasc Endovasc Surg. 1997;13:597–604
    • View In Article
    • Abstract
    • Full-Text PDF (747 KB)
    • CrossRef
17. Rutherford RB, Baker JD, Ernst C, Johnston KW, Porter JM, Ahn S, et al. Recommended standards for reports dealing with lower extremity ischemia: revised version. J Vasc Surg. 1997;26:517–538
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (1996 KB)
    • CrossRef
18. Brazier JE, Walters SJ, Nicholl JP, Kohler B. Using the SF-36 and Euro-QOL on an elderly population. Qual Life Res. 1996;5:195–204
    • View In Article
    • MEDLINE
    • CrossRef
19. Mehta T, Venkata Subramaniam A, Chetter I, McCollum P. Assessing the validity and responsiveness of disease-specific quality of life instruments in intermittent claudication. Eur J Vasc Endovasc Surg. 2006;31:46–52
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (109 KB)
    • CrossRef
20. Chetter IC, Spark JI, Scott DJ, Kent PJ, Berridge DC, Kester RC. Prospective analysis of quality of life in patients following infrainguinal reconstruction for chronic critical ischaemia. Br J Surg. 1998;85:951–955
    • View In Article
    • MEDLINE
    • CrossRef
21. Chetter IC, Spark JI, Scott DJ, Kester RC. Does angioplasty improve the quality of life for claudicants? (A prospective study). Ann Vasc Surg. 1999;13:93–103
    • View In Article
    • Abstract
    • Full-Text PDF (860 KB)
    • CrossRef