FormalPara Key Points

A structured literature review was conducted to identify randomized clinical trials comparing generalized anxiety disorder pharmacological treatments that included patient-reported outcome measures assessing health-related quality of life (HRQL), disability, functioning, or work productivity.

Results indicate that treatments that reduce anxiety symptoms are also associated with improvements in patient-reported HRQL, function, and disability.

1 Introduction

Generalized anxiety disorder (GAD) is a psychiatric disorder characterized by periods of excessive worry and anxiety occurring for more days than not over a period of at least 6 months [1]. The prevalence rate for GAD that lasts for 1 year is approximately 3 %, and the prevalence rate for a lifetime is 5 % [2]. GAD is also usually associated with other psychiatric disorders such as panic disorder, major depressive disorder or dysthymic disorder, social phobia, and specific phobia [2–4].

Along with prolonged worry and anxiety, patients with GAD may experience a variety of other symptoms, including restlessness, fatigue, difficulty concentrating, irritability, and sleep disturbances that are not due to other causes [5]. Somatic symptoms such as muscle tension and aches, sweating, dry mouth, nausea and diarrhea, trembling, feeling shaky and soreness are also common [5, 6]. The multidimensional physical and psychosocial experience of GAD contributes to substantial impairments to patients’ overall health-related quality of life (HRQL) and level of disability [2]. Patients with GAD have increased self-reported disability days and impairments in psychosocial functioning, role functioning, and work productivity [3, 7–15].

Treatments for GAD vary and include anti-depressants, anxiolytics, and cognitive behavioral therapy [16]. In the past, anti-anxiety drugs, including benzodiazepine anxiolytics, were the mainstay GAD treatment [6]. More currently, anti-depressants are considered most effective, and are first-line treatments for GAD [17]. Anti-depressants, such as selective serotonin reuptake inhibitors (SSRIs), tricyclics and monoamine oxidase inhibitors, work through various mechanisms to alleviate symptoms of GAD. SSRIs, such as fluoxetine (Prozac), sertraline (Zolfot), and escitalopram (Lexpro), work to alter the levels of serotonin within the brain. Beta-blockers, such as propanol, can also be used to treat the physical symptoms of GAD. Finally, psychotherapy and cognitive therapy are used to provide coping mechanisms for GAD patients, and have been demonstrated as helpful to those exhibiting panic, phobias, and other related symptoms [6].

Given the impact of GAD on HRQL, comprehensive evaluations of treatment for GAD must include both clinical measures [i.e., Hamilton Anxiety Rating Scale (HAM-A)] and assessments of patient-reported HRQL and functioning. The inclusion of patient-reported outcomes (PRO) that evaluate HRQL, functioning, and work productivity in the evaluation of GAD treatments provides useful information for clinicians and their patients about the benefits of treatment on patient functioning and well-being.

The purpose of this structured literature review was to further understand the impact of medication treatment on HRQL and functional outcomes. This was achieved by reviewing publications of randomized controlled trials (RCTs) comparing pharmaceutical therapies for GAD that include one or more PRO. The PRO measures reviewed include measures of HRQL, functioning, and work productivity.

2 Methods

A structured literature search was conducted to identify publications of clinical trials for GAD treatments that included PRO measures in order to evaluate the impact of GAD treatments on HRQL, disability, functioning, and work productivity. Information from the selected articles was extracted and summarized. A qualitative analysis and synthesis of the study characteristics was conducted as they related to the GAD target population.

2.1 Literature Search Strategy

The literature search was conducted in PubMed, EMBASE, Cochrane Reviews, and PsycInfo. The searches included articles published in the past 10 years in English (2004–2014; date of last search: April 7, 2014). Theoretical reviews, meta-analysis, editorials, letters, case reports, lectures, news, and comments were excluded. Reference lists of identified articles were reviewed for other articles on treatment with pharmaceutical agents and PRO measures in patients with GAD.

2.2 Identifying Health-Related Quality of Life (HRQL) Articles for Pharmaceutical Treatment in Generalized Anxiety Disorder (GAD)

The literature search included terms for GAD, HRQL, and clinical trials (i.e., “GAD” or “generalized anxiety” or “health related quality of life” or “quality of life” or “multicenter study”) and blocks of PubMed MeSh terms to elicit the most relevant articles that assess PRO instruments used in clinical trials of treatments for patients with GAD (see Appendix 1 in Electronic supplementary material). Duplicate citations from searches in the databases were identified and removed, resulting in 163 abstracts included for further review (Fig. 1). The abstract review assessed whether the publications met the following inclusion and exclusion criteria:

Fig. 1
figure 1

Summary of literature inclusion for review. GAD generalized anxiety disorder, HRQL health-related quality of life, PRO patient-reported outcome, RCT randomized controlled trial

Full size image

  • Populations under investigation were composed of patients with GAD.

  • Note: If the study included multiple types of participants (e.g., panic or other conditions with co-morbid anxiety), the publication had to report GAD results separately.

  • Studies that included a PRO measure.

  • Study designs were RCTs that included a pharmacological intervention.

  • The article did not duplicate the same clinical trial study that was reported in another article (i.e., two manuscripts reporting results from the same clinical trial).

Overall, 44 full-text articles were retrieved on the basis of the abstract review process. A second round of review was conducted in a step-wise manner on the basis of the following predetermined exclusion criteria:

  • Not an RCT (n = 13).

  • Study described alternative treatment/non-pharmaceutical interventions (note: articles that included a pharmaceutical treatment arm were included, and the data from the pharmaceutical treatment arm was summarized) (n = 4).

  • The population under investigation was not composed of patients with GAD or GAD results were not reported separately from other study populations (e.g., panic or other conditions with co-morbid anxiety) (n = 2).

  • Study did not present the results from a relevant PRO (HRQL, disability, functioning, and work productivity) (n = 6).

After the second round of review, 19 relevant articles remained. Seven of these 19 articles duplicated a clinical trial study that was reported in another article (i.e., two publications reported results from the same clinical trial). In cases where duplicate data were reported, the publication reporting the relevant PRO results was selected for review. A total of 12 articles were identified as relevant for data abstraction.

2.3 Data Abstraction

Twelve articles, representing 19 clinical trials that assessed PRO outcomes relevant to HRQL, disability, functioning, and work productivity associated with pharmaceutical therapies for GAD were reviewed. PRO variables strictly related to symptoms of anxiety were not summarized. The following variables were extracted:

  • Primary GAD inclusion/exclusion criteria

  • Sample size

  • Mean sample characteristics at baseline (age, mean age at onset of GAD, mean HAM-A total score at baseline)

  • Study design characteristics (treatment arms, trial duration, dosing schedule)

  • Relevant PRO measure results (mean change from baseline, mean difference between placebo, significance).

3 Results

3.1 Description of Trials

In total, 12 articles, representing 19 studies, were reviewed. Nine of the 12 articles presented results from a single RCT evaluating the efficacy of a pharmaceutical treatment for GAD. The remaining three articles reported results from multiple studies within each article. Demyttenaere et al. [47] was a pooled analysis of four trials evaluating the efficacy of escitalopram compared with placebo [18–20]. Endicott et al. [27] was a pooled analysis of three acute trials evaluating the efficacy of quetiapine fumarate compared with placebo [21–23]. Endicott et al. [31] reported the HRQL results of three clinical trials evaluating the efficacy of duloxetine relative to placebo [24–26] in a single article, but each trial was analyzed separately.

The clinical trials in this review included pharmaceutical treatments in the psychotropic (n = 1), anxiolytic (n = 1), and antidepressant (n = 5) therapeutic classes. In the psychotropic class, quetiapine fumarate was examined in five studies [27–29]. In the anxiolytic class, lorazepam was evaluated in one study [30]. In the antidepressant class, studies evaluating duloxetine (n = 3) [31], escitalopram (n = 2) [17], sertraline (n = 2) [32, 33], paroxetine (n = 2) [33, 34], and vortioxetine (n = 2) [35, 36] were included. The majority of the clinical trials were short term, ranging between 8 and 13 weeks. The remaining two clinical trials were long-term maintenance trials. Allgulander et al. [17] was an open-label extension trial of the treatment of escitalopram compared with placebo in the time to relapse of GAD up to 76 weeks. Katzman et al. [28] was an open-label extension trial comparing the effect of quetiapine fumarate relative to placebo in relapse prevention up to 52 weeks.

The mean HAM-A total score at baseline ranged between 20 and 27 across the clinical trials, and mean age at baseline ranged between 35 and 45 years, with the exception of Mezhebovsky et al. [37], which was a clinical trial evaluating the efficacy of quetiapine fumarate in older patients (mean age 70 years). Appendix 2 in Electronic supplementary material (Table S1) provides an overview of the study design, baseline characteristics, and HRQL efficacy results of each of the articles included in this review.

3.2 Description of PRO Measures

The PRO measures that were included in the clinical trials are presented in Table 1, and a description of the measures and evidence for the psychometric properties of the measures is provided below.

Table 1 Overview of GAD trials: PRO measures evaluated
Full size table

3.2.1 Quality of Life Enjoyment and Satisfaction Questionnaire (Q-LES-Q)

Nine of the 19 studies included the Quality of Life Enjoyment and Satisfaction Questionnaire (Q-LES-Q) [38], a self-rated questionnaire developed to assess the level of enjoyment and satisfaction experienced by patients with various psychiatric/psychological conditions in multiple areas of daily functioning. The Q-LES-Q is 83 items, however the Q-LES-Q-Short Form score may be calculated by combining the 14 items on the general activity scale with one item on satisfaction with medication and one item on global overall life satisfaction. This 16 item version of the questionnaire (Q-LES-Q-SF) is most frequently utilized in clinical trials. The Q-LES-Q has demonstrated evidence supporting reliability and validity in measuring quality of life among both psychotic and non-psychotic patients [3]. A study evaluating the measurement properties of the Q-LES-Q specifically among GAD patients demonstrated that the instrument had good reliability, validity, and responsiveness to change in this patient population [14].

3.2.2 Sheehan Disability Scale (SDS)

Seven of the 19 studies included the Sheehan Disability Scale (SDS) [39, 40], a three-item, self-rated scale developed to assess functional impairment across three domains: work, social life, and family life. The SDS has been demonstrated to be reliable and valid in primary care patients [41]. When comparing patients with and without a psychiatric disorder, patients who had a psychiatric disorder reported significantly higher levels of impairment [41]. The SDS has been able to detect treatment effects in patients with GAD [42].

3.2.3 Short Form-36 (SF-36) Health Survey

Five of the 19 studies included the Short Form-36 (SF-36), a generic health status measure utilized in a variety of therapeutic areas that measures patient health across eight domains: physical functioning, role-physical, bodily pain, general health, vitality, social functioning, role-emotional, and mental health [43]. A physical component summary score (PCS) and mental component summary score (MCS) may also be calculated.

The SF-36 can be used to provide insight into the impact of GAD on social functioning, role limitation due to emotional problems, mental health as well as physical problems related to one’s health [17]. Two of the four GAD studies that included the SF-36 only evaluated the social-functioning scale [35, 36]. One article reported all subscale scores [17], and one article reported the PCS and MCS [30].

3.2.4 EuroQol 5-Dimension (EQ-5D)

Three of the 19 studies included the EuroQol 5-Dimension (EQ-5D), another generic preference-based measure of health status used across a variety of therapeutic areas, particularly for economic evaluations. The EQ-5D measures five dimensions of health status (mobility, self-care, usual activities, pain/discomfort, and anxiety/depression) and includes a visual analog scale (EQ VAS) that generates a self-rating of HRQL (EuroQol Group [44]).

3.2.5 Endicott Work Productivity Scale

One of the articles included the Endicott Work Productivity Scale (EWPS) [32]. The EWPS is designed to assess the impact of a medical condition on work functioning through patient self-report [45]. The EWPS was developed in a sample of patients seeking treatment for depression and non-patients in the community. Psychometric testing indicated the measure was both reliable and valid in this population [45].

3.2.6 Work Limitations Questionnaire and Number of Sick Days

One of the articles evaluated the Work Limitation Questionnaire (WLQ) and the total number of sick days over a 76-week long-term maintenance period [17]. The purpose of the WLQ is to assess work limitations related to the impact of chronic health problems or treatment through patient-self report. It includes four domains: time management, physical demands, mental-interpersonal demands, and output demands. The WLQ was developed in patient samples with a variety of health conditions, including GAD, and has proven to be valid and reliable through psychometric testing [46].

3.3 Summary of Clinical Trial Findings Related to PRO

3.3.1 Q-LES-Q

Patient HRQL, as assessed by the Q-LES-Q or Q-LES-Q-SF total score, improved significantly more in patients receiving pharmaceutical treatment compared with placebo in the majority of the RCTs (Table 2). Two [25, 26] out of three studies evaluating the efficacy of duloxetine treatment over a 12- to 13-week period demonstrated a significantly greater improvement in the Q-LES-Q-SF total score in the duloxetine group compared with placebo (range 8.5–9.1 vs. 4.7–5.3; all p values < 0.001). The third study [24] did report a greater improvement in the SDS total score in the treatment group compared with placebo, but not in the Q-LES-Q-SF total score (5.7 vs. 4.2; p > 0.05).

Table 2 Summary of GAD trials: HRQL findings for pharmaceutical agents vs. placebo
Full size table

Patients receiving escitalopram (6.8 vs. 3.0; p < 0.001) [47] and sertraline (9.0 vs. 2.4; p < 0.001) [32] also improved significantly more in the Q-LES-Q total score in the treatment group as compared with placebo.

Three articles reported on the efficacy of quetiapine fumarate on Q-LES-Q-SF total score compared with placebo. One article reported the results of a pooled analysis of three clinical trials [21–23] of three different doses of quetiapine fumarate (50, 150, 300 mg/day) compared with placebo over an 8-week period [27]. The pooled analysis indicated that patients in the 150 mg/day group experienced a greater improvement in Q-LES-Q-SF total scores compared with the placebo group (11.9 vs. 8.8; p < 0.001); however, patients receiving 50 or 300 mg/day did not improve significantly more than placebo (50 mg = 9.5, 300 mg = 8.2 vs. 8.8; both p values > 0.05). A third clinical trial evaluating the efficacy of flexible-dose quetiapine fumarate over an 11-week period also detected a significantly greater improvement in the treatment group compared with placebo (14.8 vs. 4.9; p < 0.001) [29]. The long-term maintenance quetiapine fumarate trial reported that patients in the treatment group remained stable in Q-LES-Q-SF total score (−0.2), whereas patients in the placebo group declined (0.8; p < 0.05) [28].

3.3.2 SDS

The majority of the RCTs demonstrated that patients receiving pharmaceutical treatment experienced a greater improvement in overall daily life function, as assessed by the SDS total score, compared with patients in the placebo group. The SDS was included in six acute trials ranging between 8 and 13 weeks [31, 34–36]. Five of the six randomized controlled acute clinical trials demonstrated that patients treated with duloxetine (all p < 0.01 at 12–13 weeks) [24–26], paroxetine (p < 0.05 at 10 weeks) [34], and vortioxetine (p < 0.05 at 8 weeks) [35] experienced a significantly greater improvement in the SDS total score compared with participants in the placebo group (Table 2). On average, participants in these treatment groups improved between 5.8 and 8 points on the SDS, versus between 3.1 and 6.1 in the placebo groups (Table 3). A second vortioxetine acute clinical trial [36] did not detect a significant difference in SDS in the treatment groups compared with placebo (2.5 mg = −5.7, 10 mg = −5.2 vs. −4.3; both p values > 0.05), nor did it meet the primary efficacy endpoint of greater reduction in HAM-A in the vortioxetine compared with the placebo group (both p values > 0.05).

Table 3 Mean change from baseline and mean difference compared with placebo in the Sheehan Disability Scale and Q-LES-Q clinical trials of pharmaceutical agents
Full size table

The SDS was also included in a long-term study evaluating the effect of quetiapine fumarate in relapse prevention. All patients were stabilized on quetiapine fumarate in a 12- to 18-week open-label period, and then patients were randomized to receive placebo or quetiapine fumarate, and were followed for up to 52 weeks until relapse or study completion [28]. Patients in the treatment group remained relatively stable in SDS (0.1) over the 52-week relapse-prevention study period, whereas patients receiving placebo declined significantly more on average than patients in the treatment group (−1.9; p < 0.05; Table 3).

3.3.3 SF-36 Health Survey

The most commonly reported SF-36 score was the social-function subscale. Overall, evidence supported that effective pharmaceutical treatments improve social functioning in the short term, or prevent decline in the long term, as assessed by the SF-36 social-functioning scale. The SF-36 social-functioning scale was evaluated in two short-term acute clinical trials [35, 36], and one long-term maintenance study [17]. Patients receiving vortioxetine in one clinical trial [35] achieved a significantly greater improvement in SF-36 social functioning compared with placebo (26.8 vs. 18.02; p < 0.01). However, in a second clinical trial evaluating the efficacy of vortioxetine [36], no significant difference between the treatment groups compared with placebo was detected (2.5 mg/day = 18.4, 10 mg = 17.6 vs. 15.4; p > 0.05). Of note, the Bidzan et al. [35] clinical trial met the primary efficacy endpoint, whereas the Mahableshwarkar et al. [36] clinical trial did not show efficacy. In a 76-week long-term maintenance study, patients receiving escitalopram had relatively little change in social functioning over the course of the study compared with patients in the placebo group, who experienced a decline in social functioning (1.5 vs. 8.7; p < 0.01) [17].

Additional SF-36 scores were reported in three of the articles [17, 30, 34]. In a long-term maintenance study [17] evaluating the efficacy of escitalopram, patients in the treatment group declined less than the placebo group in the role-emotional (−6.5 vs. −18.2; p < 0.01) and mental-health (−1.9 vs. −10.2; p < 0.01) scales, but not in the physical-function, role-limitations, bodily pain, general health-perceptions, or vitality scale scores (all p values > 0.05). Kasper et al. [34] reported that patients taking paroxetine over 10 weeks experienced a significantly greater improvement in mental health as assessed by the SF-36 MCS compared with placebo (22.5 vs. 14.4; p < 0.01), but no significant difference in physical health as assessed by the PCS (13.1 vs. 9.2; p > 0.05). In a randomized controlled comparative effectiveness trial evaluating the efficacy of silexan and lorazepam over a 6-week period, both groups improved with treatment in the SF-36 MCS (Silexan = 21; lorazepam = 24) and the SF-36 PCS (Silexan = 12.5; lorazepam = 16.9) from baseline to week 6. No results of any statistical comparisons were reported for this study [17].

3.3.4 EQ-5D

One article [31], representing two different clinical trials [24, 25], reported on the effect of duloxetine on general health status over a 13-week treatment period, as assessed by the EQ-5D. In the first study [25], patients in the duloxetine 60-mg and 120-mg groups both experienced a greater improvement in the EQ-5D index score (60 mg = 0.19, 120 mg = 0.17 vs. placebo = 0.11; both p values < 0.05) and VAS score (60 mg = 16.7, 120 mg = 15.0 vs. placebo = 8.8; both p values < 0.05) over a 13-week treatment period compared with patients taking a placebo [25, 31]. In the Endicott et al. study [24, 31], no significant differences were detected between the treatment (duloxetine flexible dose 60–120 mg) and placebo groups in change from baseline EQ-5D index (0.10 vs. 0.05; p > 0.05) or VAS (8.0 vs. 5.4; p > 0.05) scores.

3.3.5 Work Productivity

Indices of work productivity were included in one 13-week acute RCT [32] and one 12-week open-label clinical trial with a 72-week long-term maintenance phase [17]. Participants receiving sertraline experienced a significantly greater improvement in the EWPS score relative to participants taking placebo over a 13-week period (10.1 vs. 3.9; p < 0.001) [17]. In the open-label clinical trial with a long-term maintenance phase, changes in the WLQ scores were reported for the 12-week open-label phase. Participants taking escitalopram experienced significant improvement in the WLQ index score (0.13–0.09; p < 0.001), WLQ VAS satisfaction score (6.59–8.73; p < 0.001), WLQ VAS efficacy score (6.72–8.75; p < 0.001), and the total number of sick leave days (6.7–3.9; p < 0.05). Changes in the WLQ over the long-term maintenance phase were not reported. Patients in the escitalopram group in the long-term maintenance phase did not experience any decrease in the total number of sick leave days (p > 0.05).

4 Discussion

The negative impact of GAD on patient HRQL, function, and work productivity has been well documented [3, 7–15]. A variety of anti-anxiolytics and anti-depressants are effective at reducing the severity of anxiety symptoms, as measured by the HAM-A. Given the significant impact of anxiety symptoms on functioning and well-being in patients’ lives, it is also critical to appraise the evidence of GAD treatments on patient quality of life, function, and disability. This literature review summarized the existing evidence generated in the past 10 years for the impact of pharmaceutical treatments on patient quality of life, function, and work productivity.

A total of 12 articles were identified that described the impact of a pharmaceutical treatment on quality of life, function, or work productivity in the past 10 years, representing 19 unique clinical trials of the following treatments: duloxetine, escitalopram, sertraline, paroxetine, quetiapine fumarate, and vortioxetine. With the exception of vortioxetine, multiple clinical trials have demonstrated these treatments reduce anxiety symptoms, as measured by the HAM-A. Vortioxetine, an anti-depressant approved for use in major depressive disorder, is also often used in patients with GAD; however, the results of clinical trials evaluating the efficacy in the reduction of anxiety symptoms are mixed, with one clinical trial demonstrating a significant reduction in anxiety symptoms (Bidzan et al. [35]) and a second study demonstrating no difference as compared with placebo (Mahableshwarkar et al. [36]).

Results from this literature review support that overall these treatments are effective at not only reducing anxiety symptoms, but also improving overall patient quality of life and functioning. The most frequently used PRO measure in the GAD clinical trials was the Q-LES-Q, a measure of quality of life that was developed for use across psychiatric disease indications. The measurement properties of the Q-LES-Q have been evaluated specifically in patients with GAD, and have demonstrated good reliability, validity, and responsiveness to change (Wyrwich et al. [14]). Nine out of ten trials that included the Q-LES-Q demonstrated that patients in effective treatment arms achieved a significantly different change compared with placebo. In acute clinical trials, patients in active treatment groups improved between 5.7 and 16 points on the Q-LES-Q total score. Additional research is needed to inform the interpretation of these change scores, as it is not known what amount of change in the Q-LES-Q may be considered meaningful to patients.

The SDS, a measure of functional impairment that has been used across psychiatric and non-psychiatric conditions, was also commonly included as a PRO measure in the clinical trials we reviewed. All five acute clinical trials that met their primary efficacy endpoint also demonstrated a significantly greater improvement in SDS in patients in the treatment group as compared with patients in the placebo group. These findings demonstrate that improvements in anxiety-related symptoms are associated with improvements in patient functioning.

The SF-36, a generic measure of health status used across therapeutic areas, was included in four of the 19 trials. The social-functioning subscale was the most commonly reported subscale, and results suggested that social functioning improves with treatments that significantly decrease anxiety symptoms. Three studies reported other SF-36 scale scores, and generally the results were negative for physical functioning-related scales (PCS, bodily pain, vitality, general health perceptions), but positive for mental-related scales (role-emotional, mental health). These results suggest that pharmaceutical treatments for GAD that improve anxiety symptoms may improve mental health and functioning, but with little effect on physical functioning. Alternately, these results may suggest that the physical health-related scales are not relevant domains to GAD or, rather, they are not sensitive with respect to detecting differences in this patient population.

The EQ-5D, a generic preference-based measure of health status, was included in two of the clinical trials, both of which met their primary efficacy endpoint. The EQ-5D detected a significant difference among treatment groups in one study (Koponen et al. [25]), but not in the other study (Hartford et al. [24]), suggesting that this generic measure of health status may not be sensitive with respect to detecting changes among groups in GAD trials. These findings are not surprising given that just one of five dimensions within the EQ-5D focuses on psychological symptoms.

Work productivity was only evaluated in two of the 19 clinical trials (Allgulander et al. [17, 32]), so no strong conclusions could be drawn on the effects of pharmaceutical treatments on work disability. Results from the two clinical trials that did include measures of work productivity indicate that, at least in the short-term, treatment with sertraline or escitalopram is effective with respect to improving work productivity. Additional research is needed to confirm these findings and to evaluate the relationship between improvements in anxiety-related symptoms and work productivity.

This review was limited to the past 10 years with the expectation that, given the increased focus on the value of measuring HRQL, function, and work productivity in this population in the last decade, clinical trials in the last 10 years would be most likely to include these measures. Unfortunately, in this time period, only one trial evaluating the efficacy of an anxiolytic was identified, and this was a notable limitation of the study. Further, although all of the identified published trials supported that a reduction of anxiety symptoms is associated with improvement in patient HRQL and functioning, it is possible that unpublished trials may have included PRO HRQL and functioning measures but findings were negative and therefore were not published. There were no attempts to identify unpublished data.

5 Conclusion

The results from this review suggest that reduction of anxiety symptoms is also associated with improvement in patient HRQL and functioning, particularly as it relates to mental and social functioning. Despite the demonstrated negative impact of GAD on work productivity, very few articles included measures of work productivity. Additional research is needed to evaluate the impact of GAD treatments on work functioning as this has important implications for cost-effectiveness analyses. The SDS and Q-LES-Q both demonstrate the ability to detect change among treatment groups in clinical trials that detect a treatment effect on anxiety symptoms, and may be good PRO instruments to include in clinical trials of GAD treatments. Additional research is needed to establish clinical interpretation thresholds for the meaning of change on these measures, to inform evaluation of efficacy and treatment decision making by providers and patients.