Long-term Psychosocial Outcomes of Anterior Temporal Lobectomy

Authors


Address correspondence and reprint Dr. J. E. Jones at University of Wisconsin-Madison, Department of Neurology, CG Matthews Neuropsychology Laboratory, CSC–H4/680 Mailcode 6180, 600 Highland Ave., Madison, WI 53792, U.S.A.jejones@neurology.wisc.edu

Abstract

Summary:  Purpose: To examine the long-term psychosocial outcomes of a consecutive series of patients who underwent anterior temporal lobectomy (ATL) compared with medically managed patients. This study focused primarily on actual long-term changes (mean, 5 years) in life performance. There are relatively few long-term controlled studies of actual lifetime performance changes.

Methods: The sample consisted of 61 adults who underwent ATL and 23 individuals who were evaluated for surgery but did not proceed to surgery and continued to receive medical management. Participants were interviewed 2–9 years after surgery. A structured clinical interview and review of medical records was conducted to obtain information regarding seizure frequency and psychosocial status, focusing on employment, education, driving status, financial assistance, and independent living. In addition, two questions regarding quality of life were included. Seizure frequency was evaluated for the year before the interview.

Results: Significant differences in psychosocial outcome were found between the surgery group and medical management group. Favorable psychosocial outcomes were more common among those who had surgery (46 vs. 4%). Results indicated that 68% of the surgery group exhibited improved psychosocial status compared with 5% of the medical management group. Individuals who had surgery were more likely to be driving, working full-time, living independently, and being financially independent. Remaining seizure free was not a prerequisite for improvements in psychosocial outcomes.

Conclusions: Compared with medical management, surgery has a significant positive impact on psychosocial outcomes in terms of employment, independent living, driving, and financial independence. Additionally, a person does not necessarily have to be seizure free to achieve positive changes in psychosocial outcomes.

The lack of appropriate control groups has been identified as a significant methodologic shortcoming in the literature attempting to characterize the impact of anterior temporal lobectomy (ATL) on postoperative psychosocial function. In their review of the literature, Dodrill et al. (1) reported that only one investigation to that point incorporated a control group (2). The need for appropriate controls was again reinforced recently by Lendt et al. (3), who articulated the need for controlled comparisons to characterize accurately the benefits of ATL. However, in the decade since the critical review of Dodrill et al. (1), there remain a total of only eight controlled studies (2,4–10)(Table 1).

Table 1.  Postoperative psychosocial function studies utilizing control groups
 Surgical procedureNumber of  
subjects
Outcomes
measured
Seizure free/not seizure freeFollow-up interval
  1. AED, antiepileptic drug.

Rausch & Crandall (1982)Temporal lobectomy31 surgery
9 no surgery
Dependency
Work and school
Family relationships
Nonfamily relationships
Improvement limited to SF
SF increase in IQ
1–24 mo
Guldvog et al. (1991)Anterior temporal lobectomy147 surgery
94 no surgery
Employment/school
Social pensions
Social status
Degree of social support
Dependency
AEDs
Not reported2–29 yr after surgery (mdn, 9 yr)
2–13 yr for no surgery (mdn, 9 yr)
Vickery et al. (1995)Anterior temporal lobectomy202 surgery
46 no surgery
ESI-55
Employment
Katz Adjustment Scale
Not reported5.8 yr for surgery
5.7 yr for no surgery
Baxendale & Thompson (1996)Temporal lobe surgery32 surgery
70 no surgery
Self-concept
Preoperative expectations
SF/Not SF groups did not differ significantly in item ratings1 y
Kellett et al. (1997)48% Anterior temporal lobectomy
25% Amygdalohippocampectomy
18% Temporal lesionectomy
9% Extratemporal resection
94 surgery
36 no surgery
HRQOL
Employment
Driving
SF group HRQOL scores better than those with more than 10 seizures a year1986–1994 years included in follow-up
No interval given
McLachlan et al. (1997)Temporal lobe surgery56 surgery
25 no surgery
ESI-55SF group higher
QOL at follow-up
6, 12, and 24 mo
Gilliam et al. (1999)Anterior temporal lobectomy125 surgery
71 wait-list control
ESI-55
POMS
Driving
Employment/school
AED
Seizure frequency
Adverse events profile
SF status was not associated with better HRQOL12 and 24 mo
Wiebe et al. (2001)Anterior temporal lobectomy40 surgery
40 wait-list control
LSSS
QoLIE-89
GHQ
CES-D
Not reported3, 6, 9, and 12 mo

As would be expected, there is considerable heterogeneity in the methodologic details of these eight controlled studies. For instance, Table 1 reveals that follow-up intervals varied greatly, ranging from 1 month to 29 years, with an equally wide range of psychosocial outcomes measured, which limits comparison across studies. Several studies focused on self-reported health-related quality of life [e.g., Epilepsy Surgery Inventory-55 (ESI-55)], as opposed to actual changes in life performance, although the majority of studies examined employment outcomes. Finally, these studies varied in the degree to which they attempted to identify predictors of good and poor psychosocial outcomes, including the predictive significance of seizure-free versus not seizure-free surgical outcomes.

This study examined the long-term psychosocial outcomes of a consecutive series of patients who underwent ATL compared with a group of medically managed patients. The clear focus of this study was on changes in actual life performance after surgery, including employment, independent living, driving, and financial independence. Also assessed, to a limited degree, was self-reported quality of life (11–13). The follow-up interval was long (mean, 5 years), providing a reasonable time interval with which to evaluate the eventual psychosocial outcomes of surgical intervention (10). Finally, predictors of good versus poor psychosocial outcomes were examined.

METHODS

Subjects

The subject pool consisted of a consecutive series of patients with medically intractable complex partial seizures of temporal lobe origin with or without secondary generalization who were candidates for ATL between 1990 and 1997. Participants were included if they were at least 18 years old and were of at least borderline intelligence (WAIS-R Verbal or Performance IQ >69). A total of 91 individuals met the selection criteria, and 84 (92%) participated in the follow-up study. Of the 84 participants, 61 underwent ATL, and 23 were evaluated but did not undergo surgery and served as the medical management (control) group.

Selection for ATL was determined by convergent evidence from an established surgical protocol including (a) prolonged electroencephalography (EEG) monitoring of spontaneous seizures with scalp, epidural, or subdural electrodes; (b) magnetic resonance imaging (MRI); (c) positron emission tomography (PET); (d) Wada test; and (e) neuropsychological assessment and independent speech and language evaluation. Participants in the medical management group either failed to meet criteria for surgery or declined surgery and continued to receive antiepileptic drug (AED) treatment or other standard medical treatment.

Surgery group

A total of 63 patients underwent ATL, and 61 (97%) participated in the follow-up study; one patient was lost to follow-up, and one did not respond to repeated telephone messages. Of the 61 surgical participants, 28 underwent left and 33 underwent right ATL. At follow-up, patients were 2 to 9 years after surgery (mean, 5.8 years; SD, 2.1).

Medical management group

A total of 28 patients underwent surgical evaluation but did not proceed to surgery, and 23 (82%) participated in the follow-up study. Patients did not proceed to surgery because of independent bilateral temporal lobe seizure onset (n = 15), recommendation for additional medication trials (n = 2), further surgical evaluation being refused by the patient (n = 4), or surgery being recommended but declined (n = 2). Of those who did not participate in the follow-up study, one subsequently died of a seizure, two eventually underwent surgery at other facilities, and two were lost to follow up. The patients were followed up 2 to 9 years later (mean, 5.7 years; SD, 2.0). The surgery and medical management groups did not differ in chronologic age, age at seizure onset, percentage receiving polytherapy, proportion of patients with a lesion, or time to follow-up (Table 2).

Table 2.  Baseline characteristics: surgery and medical management groups
CharacteristicSurgery
group
Medical
management
group
Age at baseline (yr) [M(SD)]31.3 (8.9)34.7 (10.3)
Age at onset (yr) [M(SD)]11.5 (10.3)15.1 (12.4)
Gender [no (%)]  
 Female28 (46%)15 (65%)
 Male33 (54%)8 (35%)
Education  
 Special education/less than high school10 (16%)2 (9%)
 GED/high school diploma33 (55%)11 (48%)
 Some college/trade certification10 (16%)9 (39%)
 College graduate5 (8%)0
 Graduate degree3 (5%)1 (4%)
Medications  
 Monotherapy31 (51%)11 (48%)
 Polytherapy35 (49%)14 (52%)
Tumor/lesion  
 Yes13 (21%)3 (13%)
 No48 (79%)20 (87%)

Psychosocial outcomes

All study participants underwent a structured clinical interview, and medical records were reviewed by an individual who was unassociated with the initial surgical and evaluation group (J.E.J.) to obtain information regarding seizure frequency and psychosocial status, with corroboration by a family member or significant other whenever difficulty recalling information was demonstrated. Assessment of psychosocial outcome focused on employment, education, driving status, financial assistance, and independent living. Information was obtained for the 1 month before the initial surgical evaluation, and long-term outcome information was garnered for the 1 month before the follow-up interview. Two questions regarding quality of life were included.

Employment

In the United States, full-time employment is defined by the Bureau of Labor Statistics (14) as working ≥35 h/week. In this study, full-time employment was defined as earning at least minimum wage and working ≥35 h/week. Part-time employment included working <35 h, unpaid homemaking, sheltered/supported employment that paid less than minimum wage, and volunteer work.

Driving

Driving was defined as holding an active driver's license and using the license actually to drive a motor vehicle. Information was obtained regarding the possible loss of license due to seizure activity at the time of the preoperative evaluation/surgery and at follow-up.

Independent living

Independent living was defined as not requiring assistance in any of three activities of daily living (e.g., organizing and taking medications, keeping appointments, and showering) and as living alone, with a roommate(s), or with significant other(s) (not if living with parents or in an assisted-living arrangement).

Financial independence

Financial independence was defined as receiving no financial assistance from any of three types of governmental assistance: Supplemental Security Income (SSI), Social Security Disability (SSDI), and W-2 (Wisconsin's AFDC replacement program). Additionally, the period of time that the individual had been receiving assistance was obtained.

Two one-item scales regarding quality of life were included. The first solicited a rating of overall quality of life on a 10-point scale, similar to questions in other instruments that evaluate health-related quality of life (15,16). The second solicited ratings on a 5-point scale; an adaptation of the Dartmouth COOP quality of life chart was used (15,17). To provide participants with an understanding of the term quality of life, the following definitions were provided: (a) a general sense of well-being related to the level of satisfaction or dissatisfaction in the areas of life that are viewed as important to you (18); and (b) the degree of satisfaction with your current life circumstances (19). As a result of the study's focus on actual life changes, a formalized quality-of-life questionnaire was not used.

Additionally, in the surgery group, participants were asked two questions regarding satisfaction with surgery. First, individuals were asked if they would have their surgery again. The second question focused on the participant's overall satisfaction with the surgery.

Individuals in the surgery and medical management groups also were classified into three composite psychosocial outcome categories for the purposes of comparison: (a) “highest status” defined by successful outcomes in all four components: full-time employment, financial independence, driving, and independent living; (b) “intermediate status” defined by successful outcomes in three but not all four categories; and (c) “lowest status” defined by success in no more than two categories.

Seizure frequency

Seizure frequency was characterized for the year before baseline evaluation and the year before the follow-up interview. Seizure frequency was categorized as follows: none, daily, weekly, monthly, and yearly (20). In addition, seizure frequency at follow-up was classified as follows: (a) seizure free, (b) auras with or without one seizure, (c) two to 12 seizures, and (d) more than 12 seizures. This system has been recommended for assessing surgery outcomes and has been found to be related to quality-of-life outcomes (21).

Neuropsychological assessment

All participants underwent baseline neuropsychological evaluation. In the surgery group, participants underwent a follow-up neuropsychological evaluation 9–12 months after ATL, and this was independent of the follow-up interview. Intelligence, memory, and emotional–behavioral status were used as potential predictors of long-term outcome. Specifically examined were Wechsler Adult Intelligence Scale (WAIS-R) Full Scale IQ and Wechsler Memory Scale Memory Quotient (WMS-MQ). In addition, from the Minnesota Multiphasic Personality Inventory (MMPI), the number of MMPI scale elevations (i.e., T-scores >70 on clinical scales 1–4 and 6–9) was used as a simple summary index to indicate psychopathology (range, 0–8). See Lezak (22) for reliability and validity information on these published measures.

RESULTS

Baseline status of surgery and medical management groups

There were no significant differences between the surgery and medical management groups in FSIQ, WMS-MQ, or the number of MMPI clinical scale elevations. The surgery group experienced more frequent complex partial seizures than the medical management group at baseline, but there was no difference in secondarily generalized seizures. Importantly, the surgery and medical management groups did not differ in baseline psychosocial status. As can be seen in Table 3, both groups exhibited relatively poor psychosocial status at baseline evaluation. Table 4 provides more detailed baseline psychosocial information. There were no differences between the surgery and medical management groups across the four psychosocial areas including employment, financial dependence, driving status, and independent living.

Table 3.  Baseline composite psychosocial status
 Surgery group
n (%)
Medical
management
group
n (%)
  1. Two cells had an expected frequency <5. χ2 (2, N = 84) = 0.668, p = 0.716.

Highest status3 (5%)2 (9%)
Intermediate status16 (26%)7 (30%)
Lowest status42 (69%)14 (61%)
Table 4.  Baseline psychosocial status: specific categories
 Surgery group
n (%)
Medical
management
group
n (%)
  1. Continuity correction used in computing all χ2 values.

Full-time employment34 (56%)11 (48%)
Living independently32 (53%)9 (39%)
Driving8 (13%)5 (22%)
Financial independence44 (72%)14 (61%)

In summary, the surgery and medical management groups appeared to be comparable in baseline seizure, cognitive, clinical, and psychosocial characteristics. The only difference concerned more frequent complex partial seizures in the surgery group.

Psychosocial outcomes for surgery and medical management groups at follow-up

Table 5 shows that there were significant group differences in psychosocial outcome [χ2 (2, N = 84) = 16.078; p < 0.001). Favorable psychosocial outcomes were significantly more common in surgical patients (46%) than medical management patients (4%), with less favorable psychosocial outcomes in surgical (24%) compared with medical management patients (65%).

Table 5. Follow-up psychosocial composite status outcomes for the surgery and medical management groups
 Surgery group
n (%)
Medical management group
(n (%)
Highest status28 (46%)1 (4%)
Intermediate status18 (30%)7 (31%)
Lowest status15 (24%)15 (65%)

Individuals in the surgery and medical management groups with less than optimal baseline psychosocial status (intact status on fewer than all four factors) were compared in regard to improvement from baseline to follow-up. Patients were considered improved if there were any gains or improvement in the number of positive psychosocial outcomes leading to a change in rating (low to intermediate or intermediate to high) from baseline to follow-up. Patients who exhibited no change or a decrease in the number of positive outcomes were classified as not improved. Results indicated that 68% in the surgery group exhibited improved psychosocial status compared with only 5% of the medical management group [χ2 (1, N = 81) = 23.18; p < 0.001; Table 6).

Table 6. Changes in presurgery to follow-up psychosocial outcome
 Surgery group
n (%)
Medical management group
n (%)
  1. Two participants from the surgery group and one from the medical management group were not included in the analysis because they were in the highest psychosocial status group at baseline. Continuity correction used in computing χ2 value.

Improvement40 (68%)1 (5%)
No improvement19 (32%)21 (96%)

Table 7 provides information regarding outcome in discrete psychosocial outcome categories and is reviewed later.

Table 7. Individual psychosocial outcomes
 Surgery group
n (%)
Medical management
group
n (%)
  1. Continuity correction used in computing all χ2 values.

Full-time employment42 (69%)9 (39%)
Living independently52 (85%)11 (48%)
Driving41 (67%)5 (22%)
Financial independence49 (80%)12 (52%)

Employment

There was a significant difference between the groups in employment outcome with significantly more in the surgery group (69%) than in the medical management group (39%) working ≥35 h/week with earnings at minimum wage or higher [χ2 (1, N = 84) = 5.00, p < 0.025]. In the surgery group, 25% were unemployed at follow-up compared with 52% in the medical management group.

In the surgery group, 74% were employed full-time preoperatively and at follow-up compared with 54% in the medical management group. At follow-up, 17% moved from part-time to full-time employment in the surgery group, as well as 15% in the medical management group. However, there was a decline among 6% of those in the surgery group who were employed full-time preoperatively and employed part-time at follow-up compared with 15% in the medical management group. Additionally, 3% were employed part-time preoperatively and at follow-up in the surgery group, as were 15% in the medical management group.

Independent living

There was a significant difference in independent-living outcomes with significantly more in the surgery group (85%) than in the medical management group (48%) who were living independently at follow-up [χ2 (1, N = 84) = 10.56; p < 0.001).

Driving

There was a significant difference between the groups in driving outcome, with significantly more in the surgery group (67%) than in the medical management group (22%) who had an active driver's license and were actually driving at follow-up [χ2 (1, N = 84 ) = 12.17; p < 0.001).

Financial independence

There was a significant difference between the groups in financial dependence, with significantly more of the surgery group (80%) than of the medical management group (52%) who were not dependent on government subsidy, and not receiving any of the three specified forms of support [χ2 (1, N = 84) = 5.32; p < 0.05). Of those in the medical management group receiving aid, 64% were receiving SSI, and 36% were receiving SSDI. In the surgery group, 28% were receiving government subsidy preoperatively compared with a decrease to 20% at follow-up. In the medical management group, 38% were receiving government subsidy at baseline compared with an increase of 48% at follow-up.

In summary, significant differences in psychosocial outcomes were found between the surgery and medical management groups. Participants who had undergone surgery were more likely to be employed full-time, living independently, driving, and not dependent on state or federal subsidies. Importantly, the groups did not differ in psychosocial status at baseline evaluation, and the outcome status lends support to the positive psychosocial impact of ATL.

Quality-of-life ratings

Participants were asked to rate quality of life at follow-up on two one-item Likert-type scales. The first (QOLI) was rated on a scale of 1 to 10 (10 was the highest rating); the second item (QOLII) rated from 1 to 5 (1 was the highest rating). The surgery group rated their quality of life at follow-up as significantly better than did medical management group on both items. Interestingly, there was only a moderate correlation found between QOLI and QOLII (r = –0.541; p ≤ 0.01), indicating that the two items may be measuring somewhat different aspects of overall quality of life.

Seizure frequency in surgery and no-surgery groups at follow-up

With the Vickery et al. (21) postoperative seizure frequency classification system, seizure frequency for the year before the follow-up interview was classified as: seizure free, auras with or without one seizure, two to 12 seizures, and >12 seizures. There was a significantly different pattern of seizure outcome between the groups [Table 8; χ2 (3, N = 84) = 42.063; p < 0.001]. As shown in Table 8, 64% of surgery participants were seizure free at follow-up compared with 4% of the medical management group.

Table 8. Seizure change at follow-up
 Surgery group
n (%)
Medical
management group
n (%)
  1. χ2 (3, N = 84) = 42.063; p < 0.001.

Seizure free39 (64%)1 (4%)
Auras with/without one seizure10 (16%)0
Two to 12 seizures6 (10%)6 (26%)
More than 12 seizures6 (10%)16 (70%)

Surgery group: psychosocial outcome as a function of seizure frequency

Table 9 shows the relation between seizure outcome and psychosocial status in the surgical participants. There was a trend toward more favorable psychosocial outcome as a function of better seizure outcome, but this relation did not reach statistical significance [χ2 (2, N = 61) = 3.489; p = 0.175].

Table 9. Follow-up seizure frequency and psychosocial status outcomes for surgery group
 Seizure free
n (%)
Not seizure free
n (%)
  1. χ2 (2, N = 61) = 3.489; p = 0.175.

Highest status21 (54%)7 (32%)
Intermediate status11 (28%)7 (32%)
Lowest status7 (18%)8 (36%)

The relationship between seizure outcome and psychosocial outcome also was examined in relation to the specific psychosocial outcome categories (Table 10). A pattern of improved psychosocial outcome with seizure-free outcome was evident for three of the four categories, but reached significance only for driving (p = 0.015).

Table 10. Follow-up status of surgery group in psychosocial outcomes for postoperative seizure free and not seizure free groups (N = 61)
 Seizure free
n (%)
Not seizure free
n (%)
  1. Continuity correction used in computing all χ2 values.

Full-time employment29 (74%)13 (59%)
Living independently34 (87%)18 (82%)
Driving31 (80%)10 (46%)
Financial independence33 (85%)16 (73%)

Finally, ratings of quality of life were examined as a function of surgical outcome. For the first quality-of-life item (QOLI), by using a scale of 1 to 10, with 1 being the lowest rating, the seizure-free group (Mdn = 9, M = 8.4, SD = 1.8) had a significantly higher QOLI rating than the not-seizure-free group (Mdn = 8, M = 7.3, SD = 1.8), t(81) = 2.33; p < 0.011. On the second item (QOLII) with a scale of 1 to 5, with 1 being the highest rating, the seizure-free group (Mdn = 1, M = 1.6, SD = 0.8) had a significantly higher quality-of-life rating than the not-seizure-free group (Mdn = 3, M = 2.7, SD = 1.1), t(81) = –1.99; p < 0.025. On both items, individuals who were seizure free rated themselves as experiencing significantly better quality of life compared with those continuing to experience seizures.

Surgery group: predictors of psychosocial outcomes at follow-up

Participants were divided into two groups based on whether they showed improved psychosocial outcome and were compared in regard to preoperative FSIQ, MQ, and MMPI scores. Significant differences were found between groups on one baseline measure, FSIQ. Individuals in both the highest and intermediate psychosocial outcome categories had higher presurgery IQ scores than did those in the lowest outcome group. The highest and intermediate categories were not different from one another.

Additionally in the surgery group, participants were separated into two groups based on improved psychosocial outcomes and were compared in regard to postoperative FSIQ, MQ, and MMPI. Only one significant difference between the psychosocial outcome groups was found. Individuals in the highest and intermediate psychosocial outcome categories had fewer postoperative MMPI clinical scale elevations >70 than did the lowest outcome group. There were no differences between the highest and intermediate categories.

Satisfaction with surgery

In the surgery group, participants were asked two questions regarding satisfaction with surgery. The first asked, “If you had to do things over, would you have your surgery again?” Only two individuals indicated that they would not have surgery again. Both individuals reported significant memory problems and frustration with the negative impact on their daily lives. The second question asked participants to use a Likert scale (10, very satisfied, to 1, very unsatisfied) to rate their overall satisfaction with surgery. A mean rating of 9.41 (SD, 1.33) indicated that the participants were overwhelmingly satisfied with the surgery. For all but two in the surgery group, participants viewed their surgical treatment as a positive treatment option to help manage their epilepsy and reduce its impact on their lives.

DISCUSSION

A surprisingly small number of investigations have examined eventual psychosocial outcomes of ATL compared with a control group consisting of medically managed patients (Table 1). The strengths of this investigation include the focus on actual life-performance changes, high retention rate at long-term follow-up (97% of surgical sample and 82% of control group), and the long-term nature of the psychosocial follow-up (5 years on average).

The groups were comparable at baseline without significant differences in regard to any of the psychosocial status categories. However, the differences between the surgical and medically managed groups in their long-term outcomes were notable. The surgery group improved significantly in terms of the proportion of patients who were able to work full-time (56% preoperatively to 69% postoperatively), live independently (53% to 85%), drive (13% to 67%), and become financially independent (72% to 80%). In contrast, patients who were managed medically did not show significant long-term changes in the proportion that were able to work full-time (48% at baseline to 39% at follow-up), live independently (39% to 48%), drive (22% to 22%), or become financially independent (61% vs. 52%).

When examined in another way (i.e., overall adequacy of psychosocial function), both groups demonstrated a very low proportion of patients functioning in the highest psychosocial category at baseline (5% surgery group, 9% medical management group), consistent with what is known about the impairing effects of intractable epilepsy. However, at follow-up, there was a significant difference between groups in the proportion of patients in the highest psychosocial category (46% surgery group, 4% medical management group). Thus surgical intervention resulted in significant longterm gains in psychosocial function whether examined in terms of individual outcomes or via composite measures.

Despite these favorable findings for surgical patients, a sizable proportion continued to exhibit considerable psychosocial difficulties at long-term follow-up. Depending on the way in which lack of psychosocial improvement was defined, from 24% to 32% of surgical patients remain impaired at follow-up (Tables 5 and 6). The percentage is still considerably lower than that observed in the medically managed patients, but reinforces the point that positive psychosocial change is not an inevitable outcome of epilepsy surgery. Epilepsy is a chronic disorder, and after living with seizures for many years, psychosocial difficulties may not automatically disappear as a result of successful surgery. This group may require more directed intervention/rehabilitation to take advantage of the reduced seizure frequency or freedom from seizures.

Although it also was clear that surgery resulted in significantly improved seizure outcomes (Table 8), the link to psychosocial outcome was not so strong as expected. In comparing seizure-free patients with those who continued to experience seizures, a higher percentage of the seizure-free patients had positive outcomes in driving, but no significant differences were found for the other psychosocial outcomes of employment, independent living, and financial independence (Table 10). Several studies have reported that freedom from seizures after surgery results in improvement in psychosocial outcomes as compared with those in individuals who have only reduced seizure frequency (2,8,23). The current study revealed mixed results. First, there were no significant differences between those who were seizure free and those who were not seizure free in terms of composite psychosocial outcomes at follow-up. However, there was a significant difference in self-reported quality of life, with the seizure-free group indicating a higher quality of life as compared with the not-seizure-free group. Improvements in psychosocial outcomes may be influenced by seizure-free status, but may not be limited to those individuals who become seizure free. It is important to identify ways in which an individual can attain optimal psychosocial outcomes with a reduction in seizure frequency, even when seizures are not completely eliminated (23,24).

The limitations of this investigation should be considered. The sample was modest in size, and patients were not randomized to the surgical or medical management arms. As controls, we used epilepsy patients who were considered for surgery but were largely not suitable for surgical intervention. However, all patients unequivocally had intractable temporal lobe epilepsy and were very comparable in baseline cognitive and psychosocial characteristics. Evaluation of outcomes was conducted through a semistructured interview developed specifically for this study, relying on self-report with confirmation by significant others as needed (25). If at all possible, information provided by the subjects was cross-checked with current medical records, but many individuals were no longer receiving medical treatment at the surgical center. Nevertheless, there is no reason to suspect that reporting biases would differ between groups when contacted by an independent investigator who was not involved in their original care or surgical-decision process.

In summary, this long-term follow-up investigation demonstrates a significant positive impact of ATL on eventual employment, independent living, driving, and financial independence. Patients did not have to be rendered completely seizure free to achieve positive changes in psychosocial outcomes. Additional studies should be conducted to determine if postsurgical rehabilitation would facilitate the attainment of optimal psychosocial outcomes.

Acknowledgment: We thank Drs. Paul Rutecki and Raj Sheth for their assistance. This work was supported in part by the Epilepsy Foundation Behavioral Sciences Postdoctoral Fellowship (J.E.J.) and by NIH (B.P.H., ROI 37738). This study was originally conducted to fulfill dissertation requirements in the Department of Rehabilitation Psychology and Special Education at the University of Wisconsin-Madison.

Ancillary