Quality of life outcomes following surgical treatment of Parkinson's disease

Quality of Life Outcomes Following Surgical Treatmentof Parkinson’s Disease

Alastair Gray, PhD,1* Iain McNamara, MB, BCh,2 Tipu Aziz, MD,3 Ralph Gregory, FRCP,4

Peter Bain, MD,5 Joanna Wilson, BA,3,6 and Richard Scott, PhD3,6

1Health Economics Research Centre, Department of Public Health, University of Oxford, Oxford, United Kingdom2Department of General Medicine, John Radcliffe Hospital, Oxford, United Kingdom3Department of Neurological Surgery, Radcliffe Infirmary, Oxford, United Kingdom

4Department of Neurology, Radcliffe Infirmary, Oxford, United Kingdom5Division of Neurosciences, Imperial College School of Medicine, London, United Kingdom

6The Russell Cairns Unit, Radcliffe Infirmary, Oxford, United Kingdom

Abstract: We assessed the impact of surgical treatment ofParkinson’s disease on quality of life using generic quality oflife instruments and utility scores. The Medical OutcomesStudy short form health survey SF-36 and Parkinson’s DiseaseQuestionnaire PDQ-39 were used before and 3–6 months aftersurgery to assess quality of life, and the results were convertedinto utility valuations. Ninety-seven patients were studied; 33underwent unilateral thalamotomy, 33 unilateral pallidotomy,20 bilateral pallidotomy, six subthalamic nucleus (STN) le-sions, four mixed lesions, and in one case bilateral STN stimu-lation. All dimensions of the SF-36 except role mental andmental health showed statistically significant improvement fol-lowing surgery. The PDQ-39 recorded significant improve-

ments in the mobility, stigma, and bodily discomfort dimen-sions. The rating scale and time trade-off scales showed statis-tically significant gains in utility of 8% and 3%, respectively.Gains were particularly marked in the bilateral pallidotomygroup. Differences in patient characteristics and selection madedirect comparisons between procedures unreliable. Quality oflife in patients with advanced Parkinson’s disease is amenableto measurement; such measurement provides tentative evidenceof significant gains in quality of life following some neurosur-gical procedures. © 2001 Movement Disorder Society.

Key Words: Parkinson’s disease; neurosurgery; SF-36;PDQ-39; utility; quality of life

The reintroduction of posteroventral medial pal-lidotomy has enjoyed renewed and widespread interestas a surgical treatment of advanced Parkinson’s dis-ease.1,2 A major reason for this has been the limitationsof current medical therapies, which lose their efficacyover time, do not prevent or significantly slow diseaseprogression, and have associated side effects (mainlymanifested as dyskinesias and psychiatric symptoms)which increase in severity and longevity with use. Im-provements in surgical procedures have led to more re-liable and reproducible results, while a deeper under-

standing of the physiology of the basal ganglia and goodmodels for the neuropathology of Parkinson’s diseasehas created new targets and strengthened the rationale forsurgical intervention.

Although a number of studies show neurological dataon the efficacy of surgical intervention, such findingshave not characteristically been in terms of health-relatedquality of life, with the capacity to compare differentsurgical procedures.3,4 As a consequence, there are im-portant gaps in knowledge concerning the nature of out-come following these procedures, which in turn has hin-dered the design of randomised controlled trials and pro-spective economic evaluations in this area. We reportdata on quality of life at baseline and follow-up on aseries of 97 patients, of whom 33 underwent unilateralthalamotomy, 33 unilateral pallidotomy, 20 bilateral pal-lidotomy, six subthalamic nucleus (STN) lesions, fourmixed lesions, and one case of bilateral STN stimulation.

*Correspondence to: Alastair Gray, Health Economics ResearchCentre, Department of Public Health, University of Oxford, Institute ofHealth Sciences, Old Road, Headington, Oxford, OX3 7LF, UnitedKingdom. E-mail: [email protected]

Received 28 February 2001; Revised 27 May 2001; Accepted 7 June2001

Published online 4 December 2001.

Movement DisordersVol. 17, No. 1, 2002, pp. 68–75© 2001 Movement Disorder SocietyPublished by Wiley-Liss, Inc.DOI 10.1002/mds.1259

68

PATIENTS AND METHODS

PatientsThe 97 patients who participated in the study were part

of a consecutive series of 126 patients undergoing sur-gery at the Radcliffe Infirmary, Oxford, or Charing CrossHospital, London, U.K., who were given the MedicalOutcomes Study short form health survey SF-36 ques-tionnaire before surgery and again 3 months after sur-gery. The 29 patients not included in the study wereexcluded for the following reasons: (1) 13 did not haveParkinson’s disease; (2) six underwent serial procedures;(3) five were not followed up or were lost to follow-up;(4) three experienced clinical events (two stroke, onepostoperative transient psychosis) between operation andfollow-up; (5) one had insufficient English to completethe questionnaires; and (6) one returned ambiguous orinconsistent data.

The baseline characteristics of the patients are shownin Table 1. Fifty-eight (60%) were men and 39 (40%)women. The mean age of patients was 65 years (range,43–83 years).

Of the 97 patients, 33 underwent unilateral Vim thal-amotomy (24 left, nine right), 33 unilateral posteroven-tral pallidotomy (24 left, nine right), 20 bilateral pal-lidotomy, four combinations of Vim and unilateral pos-teroventral pallidotomy (UPVP) lesions (on contralateralsides), six subthalamic nucleus lesions (one left, fourright, one bilateral), and one case of bilateral subthalamicnucleus stimulation. Of the 33 patients undergoing uni-lateral Vim thalamotomy, handedness was 31 right, twoleft; the mean level of tremor activity (Bain’s scale of0–105) was 7.1 (±1.1; min 5; max 9).

Patients were evaluated at baseline during clinical as-sessment prior to surgery by the same surgical team andusing the same protocol. Subsequent evaluation was per-formed at follow-up clinical assessment between 3 and 6months after surgery. Medication was only altered forpatients undergoing STN surgery. Because of differencesin patient characteristics and in selection criteria for dif-ferent procedures, direct comparisons between groups of

patients undergoing different procedures are unreliableand were not performed.

Outcome MeasuresThe two measures of quality of life administered were

the Medical Outcomes Study short form health surveySF-36, and the Parkinson’s Disease Questionnaire PDQ-39. The SF-36 has previously been shown to be relevantwhen evaluating health-related quality of life in patientswith Parkinson’s disease.6 The development and valida-tion of the Parkinson’s Disease Questionnaire has beendocumented in the UK and USA.7,8 and the instrumenthas been used widely in outcome evaluations andsurveys.3,9–11

Both the SF-36 and PDQ-39 contain a series of itemscovering eight dimensions, and each dimension is sepa-rately scored. In addition, the SF-36 contains a questionthat asks directly about the patient’s perceived change inhealth over the last 12 months. Table 2 summarises thedimensions and scoring ranges of the SF-36 and PDQ-39instruments. In addition, it is possible to derive summaryscales from each instrument. The data in the eight do-mains of the SF-36 can be used to calculate two overallsummary scores: the physical component score (PCS)and the mental component score (MCS).12 Likewise, anoverall PDQ Summary Index can also be derived fromthe data in the PDQ-39.13 Using such summary scalesmakes it possible to reduce the number of statistical com-parisons and therefore to reduce the likelihood of obtain-ing statistically significant differences due to chance.

Finally, it has recently been argued14,15 that it is pos-sible to map the descriptive health state information pro-vided by the SF-36 into preference-based utility weightswhich form the basis of quality adjusted life years(QALYs). Such preference-based measures of health sta-tus are favoured by economists when assessing the cost-effectiveness of interventions, as they can be combined

TABLE 1. Patient baseline characteristics

Characteristics Value

Sex (M/F) 58/39Age, yr (mean; range) 65; 43–83Procedures (n)

Unilateral thalamotomy (Vim) 33Unilateral posteroventral pallidotomy (UPVP) 33Bilateral posteroventral pallidotomy (BPVP) 20Contralateral combinations of Vim and UPVP 4Of which subthalamic nucleus lesions/stimulation 7Total 97

TABLE 2. Dimensions of the medical outcomes study SF-36and the Parkinson’s disease questionnaire PDQ-39

SF-36 Items PDQ-39 Items

Score, 100–0 ineach dimensiona

Score, 100–0 ineach dimensiona

Physical functioning 10 Mobility 10Social functioning 2 Activities of daily living 6Role limitations due to

physical problems4 Emotional wellbeing 6

Role limitations due toemotional problems

3 Stigma 4

Mental health 5 Social support 3Energy/vitality 4 Cognitions 4Pain 2 Communication 3General health perceptions 5 Bodily discomfort 3

a Score range from “no problem” to “maximum problem.”

QUALITY OF LIFE AND PARKINSON’S DISEASE SURGERY 69

Movement Disorders, Vol. 17, No. 1, 2002

with information on costs to calculate a cost per QALY,which in turn can be used to rank health care interven-tions. Accordingly, baseline and follow-up SF-36 datawere mapped into two utility weights, one correspondingto valuations expressed on a rating scale (respondentsassessed health states on a calibrated visual analoguescale: 0� death to 100� full health), and one in whichrespondents assessed health states using a time trade-off(TTO) procedure. In the latter case, patients had to chosebetween x years in a particular health state or y years infull health. Within this approach, the readiness to tradetime with a health problem for some lesser amount oftime in full health is interpreted as a measure of the“disutility” the respondent attaches to the health prob-lem.16 The rating scale and time trade-off values corre-sponding to particular SF-36 responses were based onresults reported from a study of 4,649 persons in a gen-eral population study who were asked to report theircurrent state of health using the SF-36, a rating scale inwhich patients indicated their current state of health as apoint between 0 and 100, and a TTO procedure (patientsasked to indicate the number of years of full health thatthey considered to be of equal value to 20 years in theircurrent health state).14 The relationship between the SF-36, rating scale, and time trade-off responses was thenestimated using regression analysis. Specifically, qualityof life scores were calculated by means of an algorithmin which coefficients from the reduced form model re-ported by Lundberg and colleagues14 corresponding tothe patient’s age, sex, and responses to specific questionsin the SF-36 were combined and expressed in the form ofa single number where 1 � full health and 0 � death.

Statistical Analysis

Student’s paired t-tests were performed to assesswhether significant differences existed between the base-line (preoperative) quality of life scores and the scores3–6 months after surgery. Comparisons were performedfor each dimension of the SF-36 and PDQ-39, as well asfor the summary scores for each instrument and for theestimated utility weights. These comparisons were madewithin each of four subgroups of patients defined by typeof procedure, and for all patients combined. Given thenumber of comparisons, differences are reported as sta-tistically significant only if P < 0.01.

Where significant differences were detected betweenbaseline and follow-up, effect sizes were also calculatedby dividing the change in the mean score between base-line and follow-up by the baseline standard deviation.Effect size is useful in giving some indication of theclinical meaningfulness of any change, with an effectsize of 1 being equivalent to a change of one standard

deviation; Kazis and colleagues17 have suggested that aneffect size of 0.2 be regarded as small, 0.5 as moderate,and 0.8 as large.

All data were analysed using SPSS v 10.

RESULTS

Table 3 gives SF-36 dimension scores at baseline andat 3–6 months after surgery, by procedure type. For pa-tients who had a unilateral thalamotomy, no significantdifferences between baseline and follow-up were de-tected in any dimension of the SF-36 other than the paindimension, where patients recorded a significant im-provement. In contrast, patients who had a unilateral pal-lidotomy recorded significant improvements in four ofthe eight dimensions of the SF-36 (physical functioning,role physical, energy/vitality, and general health percep-tions), and patients undergoing bilateral pallidotomy re-corded significant improvements in three dimensions ofthe SF-36 (social functioning, energy/vitality, and gen-eral health perceptions). The patient group having a mis-cellany of other procedures (n � 11) did not recordsignificant improvement in any dimension of the SF-36.Taking all patients together, six dimensions of the SF-36showed a statistically significant improvement followingsurgery, the exceptions being the role mental and mentalhealth dimensions.

Table 4 gives PDQ-39 dimension scores at baselineand at 3–6 months after surgery, by procedure type.Amongst unilateral thalamotomy patients, significant im-provements were recorded in the stigma and bodily dis-comfort dimensions. Unilateral pallidotomy patients re-corded improvement only in the activities of daily livingdimension. Among bilateral pallidotomy patients, sig-nificant improvements were recorded in all dimensionswith the exception of those dealing with stigma and withsocial support (i.e., problems with close relationships,lack of support from spouse, partner, friends, or family).No significant changes were observed in the group un-dergoing other miscellaneous procedures. Across all pa-tients, significant improvements were recorded in the di-mensions dealing with mobility, stigma, and bodily dis-comfort.

Table 5 shows PCS, MCS, and PDQ summary scores.The SF-36 physical component score showed significantimprovements for all groups of patients taken togetherand when grouped by procedure, while the SF-36 mentalcomponent score showed no significant changes. ThePDQ-39 summary index showed significant improve-ment gains for all patients taken together and for patientsundergoing bilateral pallidotomy, but not for other sub-groups.

A. GRAY ET AL.70


Table 6 shows the results of translating the SF-36responses into utility weights, at baseline and followingsurgery, by procedure type; in each case, the scoringrange runs from 1 � best health state to 0 � worsthealth state. The unilateral thalamotomy group showedsignificant improvements according to the rating scalebut not in the TTO results; the unilateral and bilateralpallidotomy groups registered significant improvementsusing both rating scale and time trade-off results, and the“miscellaneous procedures” group showed no significantdifferences between baseline and follow-up. Across allpatients, significant improvements in utility weightswere observed on the rating scale and the TTO results.

Table 7 gives effect sizes and change in percentagepoints over baseline for outcome measures found tochange significantly between baseline and follow-up.Clearly most effect sizes are in the range from 0.5 to 1,which could be considered moderate to large. Table 7indicates that the effect size was particularly large in thegroup undergoing bilateral pallidotomy: of 13 measuresor dimensions of quality of life that showed statisticallysignificant improvement, eight displayed an effect sizegreater than 1 and none displayed an effect size smallerthan 0.6.

DISCUSSION

We report on measures of health-related quality of lifebefore and after neurosurgical treatment of Parkinson’sDisease in a consecutive series of mixed procedures.Baseline and follow-up assessments were carried outwith the SF-36 and PDQ-39, and the SF-36 was alsomapped into utility weights that could facilitate qualityof life calculations for economic evaluations.

The results clearly indicate that patients undergoingunilateral pallidotomy and particularly, bilateral pal-lidotomy, recorded statistically significant improvementsacross a wide range of dimensions of quality of life, witheffect sizes that could generally be considered moderateto large. The subgroup of patients receiving unilateralthalamotomy demonstrate significant improvements inrelatively few domains of quality of life following sur-gery. The patients in the “miscellaneous procedures” cat-egory showed no statistically significant gains in anymeasure of quality of life, but the small size (n � 11)and heterogeneity of this subgroup must be borne inmind.

The overall change in quality of life indicated by thetwo utility measures (rating scale and TTO) was statis-tically significant for all patients, with an improvement

TABLE 3. SF-36 dimension scores at baseline and at 3 months after surgery, byprocedure type

UT(n � 33)

UP(n � 33)

BP(n � 20)

Misc other(n � 11)

All(n � 97)

Physical FunctioningBaseline 54 (26) 28 (23) 24 (18) 33 (18) 37 (26)3 months 55 (29) 40 (27)** 40 (29) 35 (25) 44 (29)**

Role-physicalBaseline 25 (37) 11 (18) 11 (25) 18 (31) 16 (29)3 months 39 (41) 30 (36)** 24 (36) 13 (32) 30 (38)***

Role-mentalBaseline 57 (47) 53 (43) 62 (39) 67 (35) 57 (42)3 months 61 (45) 63 (39) 70 (42) 60 (41) 63 (41)

Social functioningBaseline 55 (10) 43 (17) 46 (13) 52 (25) 49 (16)3 months 65 (31) 54 (32) 67 (29)** 43 (29) 59 (31)**

Mental HealthBaseline 69 (18) 61 (20) 70 (17) 60 (17) 66 (19)3 months 70 (21) 67 (20) 77 (17) 63 (16) 69 (19)

Energy/vitalityBaseline 49 (16) 37 (20) 37 (15) 31 (17) 40 (18)3 months 47 (23) 50 (20)*** 52 (16)** 36 (16) 48 (20)***

PainBaseline 34 (28) 51 (29) 45 (27) 48 (35) 44 (29)3 months 73 (21)*** 59 (33) 69 (29) 69 (33) 67 (29)***

General Health PerceptionsBaseline 58 (19) 43 (21) 41 (18) 50 (24) 49 (21)3 months 60 (18) 59 (23)** 56 (20)** 50 (24) 57 (21)***

Values expressed as mean (S.D.).**P < 0.01 ***P < 0.001UT, unilateral thalamotomy; UP, unilateral pallidotomy; BP, bilateral pallidotomy.



of eight percentage points (from 0.56 to 0.64) in therating scale, and of three percentage points (from 0.66 to0.69) in the TTO scale. The gains were particularlymarked in the bilateral pallidotomy group (12 percentagepoints in the rating scale and seven percentage points inthe TTO).

The differences in outcome between unilateral thala-motomy and pallidotomy groups is perhaps not unex-pected given the clear differences in disease severity and

the nature of the presenting symptomatology (see Tables3 and 4). Patient selection criteria for thalamotomy con-fines the presenting symptomatology to unilateral restingtremor; although there was also not infrequently someassociated ipsilateral rigidity and pain/discomfort. Onewould not therefore anticipate postoperative improve-ments other than in the dimensions where it was in factobserved: PDQ-39 stigma and bodily discomfort and SF-36 pain. Indeed patients’ preoperative presentation rou-

TABLE 4. PDQ-39 scores at baseline and 3 months after surgery, by procedure type

PDQ-39 DimensionsUT

(n � 33)UP

(n � 33)BP

(n � 20)Misc other(n � 11)

All(n � 97)

MobilityBaseline 31 (28) 69 (24) 76 (20) 69 (19) 58 (31)3 months 32 (34) 61 (31) 51 (28)** 57 (35) 49 (34)***

Activities of daily livingBaseline 46 (26) 54 (24) 60 (16) 64 (27) 54 (24)3 months 33 (33) 38 (29)*** 33 (25)*** 51 (30) 36 (29)

Emotional well beingBaseline 25 (20) 33 (21) 34 (21) 48 (24) 32 (21)3 months 22 (21) 31 (23) 22 (23)** 40 (25) 27 (23)

StigmaBaseline 48 (29) 41 (27) 37 (31) 65 (26) 45 (29)3 months 28 (29)** 30 (32) 19 (31) 47 (29) 29 (31)***

Social supportBaseline 8 (17) 20 (22) 14 (24) 17 (20) 15 (21)3 months 11 (21) 14 (20) 14 (24) 24 (36) 14 (23)

CognitionsBaseline 25 (16) 29 (20) 32 (16) 30 (25) 29 (18)3 months 24 (19) 27 (23) 19 (13)** 35 (26) 25 (20)

CommunicationBaseline 17 (20) 35 (25) 40 (24) 21 (20) 29 (24)3 months 19 (20) 32 (29) 24 (19)** 49 (27) 27 (25)

Bodily discomfortBaseline 41 (23) 53 (26) 59 (19) 39 (23) 49 (24)3 months 25 (21)*** 44 (31) 33 (22)*** 38 (36) 34 (27)***

Values expressed as mean (S.D).**P < 0.01 ***P < 0.001UT, unilateral thalamotomy; UP, unilateral pallidotomy; BP, bilateral pallidotomy.

TABLE 5. SF-36 physical component summary and mental component summary and PDQ-39summary index, at baseline and at 3 months after surgery, by procedure type

UT(n � 33)

UP(n � 33)

BP(n � 20)


All(n � 97)

SF-36Physical component score

Baseline 26 (10) 21 (8) 17 (7) 23 (7) 22 (9)3 months 35 (11)*** 28 (14)** 27 (12)** 26 (16) 30 (13)***

Mental component scoreBaseline 48 (13) 44 (11) 49 (10) 47 (10) 47 (11)3 months 46 (13) 48 (10) 53 (12) 42 (7) 48 (11)

PDQ-39Summary index

Baseline 30 (15) 42 (15) 44 (11) 44 (15) 39 (15)3 months 24 (17) 35 (21) 27 (16)*** 43 (22) 30 (19)***


A. GRAY ET AL.72


tinely included salient concerns about the acute embar-rassment of tremor in public. A more global functionalimprovement might only be predicted if the tremor wasof the action/intention type.

Pallidotomy, on the other hand, was selected for pa-tients with more advanced and severe presenting symp-toms, including dyskinesia, bradykinesia, and (moremarked) rigidity and pain, and these more routinely ex-

tended into the lower body. Clearly in this group, thefunctional impact of symptoms upon general physicalfunctioning (e.g., mobility) and activities of daily livingwould be much greater than in the thalamotomy group.These differences are reflected in this group’s preopera-tive presentation (see Tables 3 and 4) and, broadly, in thedomains where postoperative gains were observed. Inone sense, the absence of any significant postoperative

TABLE 6. Estimated quality of life, derived from rating scale and time trade-off valuations ofSF-36 responses, at baseline and 3 months after surgery, by procedure

UT(n � 33)

UP(n � 33)

BP(n � 20)


All(n � 97)

Rating scaleBaseline 0.59 (0.07) 0.54 (0.08) 0.52 (0.08) 0.57 (0.09) 0.56 (0.08)3 months 0.67 (0.13)*** 0.62 (0.13)*** 0.64 (0.11)*** 0.58 (0.11) 0.64 (0.13)***

Time trade-offBaseline 0.69 (0.08) 0.63 (0.09) 0.62 (0.10) 0.69 (0.07) 0.66 (0.09)3 months 0.70 (0.10) 0.68 (0.11)** 0.69 (0.12)** 0.67 (0.09) 0.69 (0.11)**


TABLE 7. Effect size* and percentage points of change in SF-36 dimensions, PDQ-39 dimensions, SF-36 physical componentsummary and mental component summary, PDQ-39 summary index, rating scale, and time trade-off score, by procedure type

UT(n � 33)

UP(n � 33)

BP(n � 20)


All(n � 97)

Effectsize

% Pointchange

Effectsize

% Pointchange

Effectsize

% Pointchange

Effectsize

% Pointchange

Effectsize

% Pointchange

SF-36Physical

Functioning 0.5 11 0.3 8Role-physical 1.1 19 0.5 13Role-mentalSocial functioning 1.6 21 0.7 11Mental Health 0.2 4Energy/vitality 0.7 13 1.0 15 0.4 7Pain 1.4 39 0.8 23General

HealthPerceptions 0.8 16 0.8 15 0.4 9

PDQ-39Mobility 1.3 25 0.3 9Activities of daily living 0.7 16 1.7 28Emotional well being 0.6 12Stigma 0.7 20 0.6 16Social supportCognitions 0.8 13Communication 0.7 16Bodily discomfort 0.7 16 1.4 26 0.6 15

Rating scale 1.2 8 1.0 8 1.5 12 1.0 8TTO score 0.6 1 0.7 7 0.4 4SF-36

Physical component score 1.0 10 0.9 7 1.4 10 0.9 8Mental component score

PDQ-39Single Index 1.6 17 0.6 8

* Statistically significant change observed from baseline to follow-up.UT, unilateral thalamotomy; UP, unilateral pallidotomy; BP, bilateral pallidotomy; TTO, time trade-off.



change in more global functional domains followingthalamotomy to some extent serves to validate thechanges observed on the SF-36 and PDQ-39 followingpallidotomy, for it suggests that these generic anddisease-specific questionnaires are appropriately sensi-tive, and that the gains observed are unlikely to reflect aplacebo effect.

Although the results reported here are suggestive ofdifferences between groups and procedures, these shouldbe viewed with great caution. Comparisons of functionalgain between the unilateral and bilateral groups couldonly be made if patients were randomly assigned to oneof the two procedures where the unilaterals initially pre-sented as hemi-parkinsonian. Furthermore, patients un-dergoing thalamotomy are tremor-predominant, whilethose undergoing pallidotomy are dyskinesia-predomi-nant, and unilateral pallidotomy patients are less disabledthan bilateral pallidotomy patients preoperatively andtherefore cannot be matched. Finally, no sufficientlylarge group of patients receiving deep brain stimulationwas available for assessment, and the impact of this pro-cedure on health-related quality of life must also be con-sidered in future research. The purpose of this study isnot to compare the effectiveness of one procedure withanother; rather, to explore the utility associated withquality of life measures in this setting, to give some ideaof size of any improvements, and help with the planningand design of future randomised controlled trials.

A number of important aspects of these results shouldbe noted. First, as Lundberg and colleagues14 acknowl-edge, the algorithm used to convert SF-36 responses intoutility weights still requires further development work.For example, the TTO results on which the algorithm isbased were obtained from postal questionnaires, butface-to-face survey is considered more reliable. Thepopulation on which the algorithm is based was alsoconsiderably healthier than the patients considered here;in the population survey, the mean health state utility was0.83 using the rating scale, and 0.90 using the time trade-off; in patients with Parkinson’s disease, baseline meanswere 0.56 and 0.66, respectively.

The health status of patients in this study is also sub-stantially lower on average than in representative generalpopulation samples. For example, the mean SF-36 physi-cal component score for all patients prior to surgery was22, compared with a mean of 45 amongst men andwomen in the U.K. general population between the agesof 55 and 64 years.18

Personal clinical experience (T.A.) with Parkinson’sdisease patients suggests that some place a valuation on

their health states preoperatively that is worse than death,and express preferences for a failed operation and deathover remaining in their current health state. This extremeview is not illustrated in the preoperative TTO scores,which are in a relatively narrow range across all patients(from 0.44 to 0.85).

One way of potentially avoiding some of these prob-lems would be to record directly the quality of life ofpatients undergoing surgery using a generic descriptiveinstrument such as the Euroqol EQ-5D or Health UtilitiesIndex III.19,20 These instruments are accompanied by ref-erence weights, or “tariff” values, for each of the de-scriptive health states they contain, and would thereforedispense with the need to map results from the SF-36using an algorithm obtained from another study. Someresults from community surveys of patients with Parkin-son’s disease indicate that this is a feasible approach,21

and prospective collection of such data is now underway. Another approach would be to explore preferencesconcerning risk, outcome, and health states directly withpatients.

Our results are based on quality of life changes at 3–6months after surgery. The effectiveness and ultimatelythe cost-effectiveness of surgical intervention will re-quire longer-term evidence of effectiveness. Some stud-ies have reported benefits from surgery in specificgroups of patients over longer periods of time. For ex-ample, Fine and colleagues4 have reported that improve-ments on the Unified Parkinson’s Disease Rating Scale(UPDRS) amongst patients undergoing unilateral pal-lidotomy were sustained over periods up to 5 1/2 years.Outcome in terms of quality of life questionnaire mea-sures and neurological rating scales over longer follow-up periods for the group of patients reported here arecurrently being prepared by our group.

In addition to the amount and duration of benefit, fullevaluation of surgery for Parkinson’s disease requiresassessment of what would have happened in the absenceof surgery rather than a straight before and after surgerycomparison. Insofar as the health of patients undergoingsurgery would otherwise have deteriorated, the improve-ments reported here in comparison with presurgicalhealth levels may understate the true gains from surgery.Such questions are most likely to be answered by largerandomised controlled trials within which quality of lifeand resource use data can be collected prospectively.Until such evidence is available, however, our findingsindicate that quality of life in patients with advancedParkinson’s disease is amenable to measurement, andprovide tentative evidence of significant gains in qualityof life following some neurosurgical procedures.

A. GRAY ET AL.74


Acknowledgments: Tipu Aziz is supported by the MedicalResearch Council. We are grateful to Crispin Jenkinson andEmma McIntosh for comments and advice on an earlier versionof this manuscript.

REFERENCES

1. Laitinen LV, Bergenheim AT, Hariz MI. Leksell’s posteroventralpallidotomy in the treatment of Parkinson’s disease. J Neurosurg1992;76:53–61.

2. Lang AE. Surgery for levodopa-induced dyskinesias. Ann Neurol2000;47:S193–S199.

3. Martinez MP, Valldeoriola F, Molinuevo JL, Nobbe FA, Rumia J,Tolosa E. Pallidotomy and quality of life in patients with Parkin-son’s disease: an early study. Mov Disord 2000;15:65–70.

4. Fine J, Duff J, Chen R, Chir B, Hutchison W, Lozano AM, LangAE. Long-term follow-up of unilateral pallidotomy in advancedParkinson’s disease. N Engl J Med 2000;342:1708–1714.

5. Bain PG, Findley LJ, Atchison P, Behari M, Vidailhet M, GrestyM, Rothwell JC, Thompson PD, Marsden CD. Assessing tremorseverity. J Neurol Neurosurg Psychiatry 1993;56:868–873.

6. Jenkinson C, Peto V, Fitzpatrick R, Greenhall R, Hyman N. Self-reported functioning and well-being in patients with Parkinson’sdisease: comparison of the short-form health survey (SF-36) andthe Parkinson’s Disease Questionnaire (PDQ-39). Age Ageing1995;24:505–509.

7. Peto V, Jenkinson C, Fitzpatrick R. PDQ-39: a review of the de-velopment, validation and application of a Parkinson’s diseasequality of life questionnaire and its associated measures. J Neurol1998;245 Suppl 1:S10–S14.

8. Bushnell DM, Martin ML. Quality of life and Parkinson’s disease:translation and validation of the US Parkinson’s Disease Question-naire (PDQ-39). Qual Life Res 1999;8:345–350.

9. Straits TK, Fields JA, Wilkinson SB, Pahwa R, Lyons KE, KollerWC, Troster AI. Health-related quality of life in Parkinson’s dis-ease after pallidotomy and deep brain stimulation. Brain Cogn2000;42:399–416.

10. Peto V, Fitzpatrick R, Jenkinson C. Self-reported health status andaccess to health services in a community sample with Parkinson’sdisease. Disabil Rehabil 1997;19:97–103.

11. Harrison JE, Preston S, Blunt SB. Measuring symptom change inpatients with Parkinson’s disease. Age Ageing 2000;29:41–45.

12. Ware-JE J, Kosinski M, Bayliss MS, McHorney CA, Rogers WH,Raczek A. Comparison of methods for the scoring and statisticalanalysis of SF-36 health profile and summary measures: summaryof results from the Medical Outcomes Study. Med Care 1995;33(4Suppl):AS264–AS279.

13. Jenkinson C, Fitzpatrick R, Peto V, Greenhall R, Hyman N. TheParkinson’s Disease Questionnaire (PDQ-39): development andvalidation of a Parkinson’s disease summary index score. AgeAgeing 1997;26:353–357.

14. Lundberg L, Johannesson M, Isacson DG, Borgquist L. The rela-tionship between health-state utilities and the SF-12 in a generalpopulation. Med Decis Making 1999;19:128–140.

15. Brazier J, Usherwood T, Harper R, Thomas K. Deriving apreference-based single index from the UK SF-36 Health Survey.J Clin Epidemiol 1998;51:1115–1128.

16. Torrance GW. Measurement of health state utilities for economicappraisal: a review. J Health Economics 1986;5:1–30.

17. Kazis LE, Anderson JJ, Meenan RF. Effect sizes for interpretingchanges in health status. Med Care 1989;27:S178–S189.

18. Jenkinson C. The SF-36 physical and mental health summary mea-sures: An example of how to interpret scores. J Health Ser ResPolicy 1998;3:92–96.

19. EuroQol Group. EuroQol—a new facility for the measurement ofhealth-related quality of life. Health Policy 1990;16:199–208.

20. Torrance GW, Feeny DH, Furlong WJ, Barr RD, Zhang Y, WangQ. Multiattribute utility function for a comprehensive health statusclassification system. Health Utilities Index Mark 2. Med Care1996;34:702–722.

21. Schrag A, Selai C, Jahanshahi M, Quinn NP. The EQ-5D—a ge-neric quality of life measure-is a useful instrument to measurequality of life in patients with Parkinson’s disease. J Neurol Neu-rosurg Psychiatry 2000;69:67–73.



Documents

Quality of life outcomes following surgical treatment of Parkinson's disease