Psychometric and Clinical Evaluation of the Clinician (VQIDS-C₅) and Self-Report (VQIDS-SR₅) Versions of the Very Quick Inventory of Depressive Symptoms

Plain Language Summary

A short questionnaire to measure depressive symptoms would be easier for patients to take either on paper or by smartphone and provide a quick status check to their clinicians. The authors developed the 5-item Very Quick Inventory of Depressive Symptomatology patient self-rating (VQIDS-SR₅) and clinician-rating (VQIDS-C₅) using five questions from the 16-item QIDS₁₆. This study was conducted to determine whether these measures are accurate and suitable for the purposes mentioned above.

This study used data from outpatients 18–75 years of age with major depressive disorder treated for up to 14 weeks with citalopram and who completed measures of depressive symptoms, including the QIDS₁₆ versions at baseline and each treatment visit. The researchers extracted the VQIDS₅ items from the completed QIDS₁₆ versions and compared the information to that from the full QIDS₁₆ versions and other depression rating scales.

Findings indicated that both the self-report and clinician versions of the VQIDS possess nearly identical score ranges for categories of depressive symptom severity (mild, moderate, etc) and treatment outcomes (response, remission, etc.). Both VQIDS detect symptom changes as well as the 16-item QIDS₁₆.

The VQIDS₅ versions are suitable for rating patient depressive symptoms and for remote monitoring or telepsychiatry visits. Patients could complete the VQIDS-SR5 on their smartphone to monitor their depression and to provide that information to their clinicians. How well the VQIDS₅ assesses depression in juveniles or the elderly, and how it compares to other symptom severity ratings deserves study.

Introduction

Measurement-Based Care (MBC) improves the outcomes of patients with depression,¹ but its implementation faces challenges,² including changing the workflow and practice management habits of clinicians and fears of increased time requirements. On the other hand, patients and/families are often eager to participate in their own care. If armed with appropriate information and easy-to-use measurement tools, they often undertake habit change and increase participation in their own care.

Presently, the self-reported 9-item Patient Health Questionnaire (PHQ-9) is a widely used tool for assessing depressive symptom severity and treatment outcome.³ The PHQ-9 was devised as a screening tool for a major depressive episode (MDE) based on the Diagnostic and Statistical Manual IV.⁴ The PHQ-9 uses a 2-week observation period and asks about the pervasiveness rather than the severity of depressive symptoms (eg, “over the last 2 weeks, how much have you been bothered by any of the following problems?”). Criterion depressive symptoms do wax and wane in severity. Thus, a patient with severe insomnia occurring only on a few nights/week might find the rating to be challenging. The 2-week assessment period could also make detection of change more sluggish than scales that use shorter assessment periods. While the PHQ-9 does assess all symptomatic criteria that define an MDE, its nine items may make administration by smartphone longer than is necessary to gauge the core symptoms of depression, dysphoria and anhedonia.

To develop a more user-friendly depressive symptom rating tool for potential smartphone use, we created the Very Quick Inventory of Depressive Symptomatology (VQIDS₅).⁵ The VQIDS₅ includes 5 items from the 16-item Quick Inventory of Depressive Symptomatology (QIDS₁₆)⁶ that best approximate the 6 items on Bech’s 6-item Hamilton Rating Scale for Depression (HRSD₆).⁷ A variety of clinical trials with depressed outpatients found that the HRSD₆ total score accounted for more than half of the variance in overall depressive symptom outcome provided by the 17-item HRSD (HRSD₁₇) total score,^8–10 thereby providing a briefer assessment of depressive symptom outcome. The resulting VQIDS self-report (VQIDS-SR₅) and clinician-rated (VQIDS-C₅) include sad mood, self-outlook, involvement, fatigue, and psychomotor slowing. Unlike the HRSD₆, however, neither version of the VQIDS₅ nor the original QIDS₁₆ ratings included an anxiety item. Anxiety, however, is not a criterion symptom of an MDE, and that item had the smallest loading in the principal component analysis⁸ of the HRSD₆, likely because some MDEs are characterized by high anxiety and others are not.¹¹

Initially, we evaluated both the clinician-rated and self-report versions of the VQIDS in a sample of outpatients with chronic or recurrent MDD being treated in psychiatric and primary care settings in the Combining Medications to Enhance Depression Outcomes (CO-MED) study.¹² The VQIDS-SR₅ and the VQIDS-C₅ were unifactorial and were at least as sensitive to change as the longer 16-item QIDS-SR₁₆ and QIDS-C₁₆ ratings. The need for replication in an independent sample was recognized.⁵

The current report further evaluated the psychometric features of the VQIDS-SR₅ and VQIDS-C₅; created crosswalks between their total scores and those of the QIDS-SR₁₆, QIDS-C₁₆ and HRSD₆; established symptom severity (eg, none, mild, moderate, etc) and recommended various outcome categories (eg, no benefit, partial response, response without remission, and remission). We also compared the proportions of patients that fell into these various outcome categories based on the VQIDS ratings with those that fell into those categories based on the QIDS-SR₁₆, QIDS-C₁₆ and HRSD₆.

Materials and Methods

This report employed data obtained from the first treatment step in the Sequenced Treatment Alternatives to Relieve Depression (STAR*D) trial (NCT00021528). The STAR*D protocol and procedures were approved by the Institutional Review Board at the University of Texas Southwestern Medical Center (National Coordinating Center), the University of Pittsburgh (data coordinating center), and each participating center and clinic, according to the Declaration of Helsinki. All participants provided written informed consent before data collection and trial entry. The rationale, methods, and design of STAR*D are detailed elsewhere.¹³

Overall, 14 regional Centers across the United States oversaw protocol implementation at public or private sector clinical sites that provide primary (N=18) or psychiatric (N=23) outpatient care. This was an open treatment, rater blinded, multistep treatment trial that used an equipoise stratified randomization.¹³

Participants

Outpatients (18–75 years of age) with single or recurrent nonpsychotic major depressive disorder who were seeking care (as opposed to enrolling symptomatic volunteers) were eligible if their baseline HRSD₁₇ total score was at least 14 and their clinicians decided that outpatient antidepressant medication was appropriate.

Diagnostic and Outcome Measures

The diagnosis of nonpsychotic major depressive disorder was made by the clinician but confirmed using a DSM-IV criterion symptom checklist. An initial HRSD₁₇, QIDS-C₁₆ and QIDS-SR₁₆ were collected. Telephone interviews were conducted to obtain the HRSD₁₇, which was the primary outcome. A missing exit HRSD₁₇ was declared to be non-remission. Secondary outcomes included the QIDS-SR₁₆ and QIDS-C₁₆ collected at baseline and at each treatment visit.

This report used the HRSD₆ (derived from its parent scale, the HRSD₁₇), the VQIDS-C₅ and VQIDS-SR₅ (derived from the QIDS-C₁₆ and QIDS-SR₁₆, respectively), and the 30-item Inventory of Depressive Symptomatology (IDS-C₃₀)¹⁴ as outcome measures. Details and psychometric properties on each of the scales can be found elsewhere.^{5–7,10,14,15}

Treatment

In this first treatment step, once-a-day citalopram was prescribed with the aim of reaching symptom remission, defined as a QIDS-C₁₆ score ≤5 at a maximum dose of 60mg/d. Dosing recommendations were flexible and based on MBC procedures as informed by the Frequency, Intensity, and Burden of Side Effects Ratings (FIBSER)¹⁶ and the QIDS-C₁₆ scores at each treatment visit.

The protocol recommended treatment visits at 2, 4, 6, 9, and 12 weeks with an optional week 14 visit if needed. After an optimal trial, remitters and responders could enter the 12-month naturalistic follow-up, though all responders who did not achieve remission were encouraged to enter the second treatment step. Participants could discontinue citalopram before 12-weeks if 1) they experienced intolerable side effects, 2) an optimal dose increase was not possible due to side effects or participant choice, or 3) significant symptoms (QIDS-C₁₆ score ≥9) were present after 9 weeks at maximally tolerated doses. Patients could opt to move to the next treatment level if they had intolerable side effects or if the QIDS-C₁₆ score was >5 after an adequate trial in terms of dose and duration.

Concomitant Medications

Concomitant treatments for current general medical conditions and for citalopram side effects were permitted based on clinical judgment. Stimulants, anticonvulsants, antipsychotics, alprazolam, non-protocol antidepressants (except trazodone ≤200 mg at bedtime for insomnia), and depression-targeted psychotherapies were proscribed.

Statistical Analysis

The VQIDS-C₅ and VQIDS-SR₅ total scores were computed from the appropriate subset of items of the QIDS-C₁₆ and QIDS-SR₁₆, respectively. Also, the HRSD₆ was taken from items of the HRSD₁₇. To be included in an analysis, participants must have had all items for all scales used in the analysis. The QIDS-C₁₆ and QIDS-SR₁₆ were collected at intervals from baseline to the end of Level 1. Visit windows, which extend from 1 week before to 1 week after weeks 2, 4, and 6, were created based on visit date. The HRSD₁₇ and IDS-C₃₀ were collected only at baseline and exit from this treatment step. Exit analyses included the HRSD₁₇ and IDS-C₃₀ at exit and the last available QIDS-C₁₆ and QIDS-SR₁₆ up to week 14.

Internal consistency of the VQIDS-C₅ and VQIDS-SR₅ was assessed using Cronbach’s alpha¹⁷ and item-total correlations (corrected for item/total overlap) using Spearman correlations¹⁸ at the exit assessment.

Item Response Theory (IRT)¹⁹ methods were implemented in MULTILOG²⁰ using Samejima’s graded IRT model.²¹ The assumption of uni-dimensionality necessary for IRT methods was determined by principal component analysis with the number of dimensions determined by parallel analysis where dimensionality is assessed by comparing eigenvalues from the real dataset to eigenvalues from simulated datasets with randomly generated uncorrelated variables.²² See Supplement for details. Exit measurements were used for the QIDS-C₁₆, QIDS-SR₁₆, VQIDS-C₅, VQIDS-SR₅, and HRSD₆.

IRT analyses were used to compare the remaining measures using the test information function, which enables one to compare the precision of two or more measures across all levels of symptom severity. Also, we created tables that enable conversion of VQIDS-C₅ and VQIDS-SR₅ total scores into total scores on the other measures with equivalent levels of depression using the exit assessment. The procedure of Orlando et al²³ (and associated software) was used to derive an IRT score (measure of depression scaled to a mean 0 and standard deviation 1) for each possible total score on each of the scales. Item scores with the same IRT score were then equated.

The percent of participants with various levels of improvement from baseline to weeks 2, 4, 6, and exit were reported. The levels of improvement were applied to all scales and were as follows: no meaningful benefit (≤35% improvement), partial response (>35% improvement and <50% improvement), response (≥50% improvement) without remission, and remission HRSD₆ ≤3, VQIDS-C₅ ≤2, VQIDS-SR₅ ≤2). These levels of improvement in depressive symptoms were based on established clinical consensus²⁴ and empirical reports (see for example Conway et al²⁵ and Dunlop et al²⁶). The numerical remission thresholds for each measure were similarly based on clinical convention (eg, HRSD₁₇ ≤7; QIDS-C₁₆ ≤5; QIDS-SR₁₆ ≤5)²⁷ or on the equivalent values arrived at by IRT for the brief versions (eg, HRSD₆ ≤3, VQIDS-C₅ ≤2, VQIDS-SR₅ ≤2). The strength of agreement between each pair of measures was assessed by kappa statistic for remission vs no remission, response (with or without remission) vs no response, and benefit (over 35% improvement) vs no benefit 35% or less improvement). Sensitivity of the measures to change over time was assessed by computing the percent change from baseline to weeks 2, 4, 6, and exit. Differences in percent change between each pair of measures was tested using paired t-test.²⁸

Results

In the STAR*D trial, 4045 participants enrolled into Level 1 and had some baseline data. The number of participants with all QIDS₁₆, HRSD₁₇, and IDS-C₃₀ items at exit was 2493, while the number of participants with all items at both baseline and exit was 2352. The number of participants with all QIDS-C₁₆ and QIDS-SR₁₆ items for weeks 2, 4, and 6 were 2732, 2273, and 1984, respectively.

CTT Properties of VQIDS-C₅ and VQIDS-SR₅

Cronbach’s alphas (based on last visit data during citalopram treatment, which could have been up to 14 weeks following baseline) were 0.86 for the VQIDS-C₅ and the VQID-SR₅; and 0.87 for the QIDS-C₁₆ and the QIDS-SR₁₆. Corresponding Cronbach’s alpha values were 0.87 for the HRSD₁₇, 0.86 for the HRSD₆ and 0.92 for the IDS-C₃₀. All values fell within an acceptable range.¹⁷

Table 1 summarizes the Spearman item-total correlations computed at exit. To avoid inflating these correlations, each specific item was removed from the total score when its item total correlation was calculated. For example, if sad mood was included in the total score, the correlation of sad mood with the total would be artificially inflated. To avoid this, sad mood was removed from the total.

Table 1 Item-Total Spearman Correlationsa Computed at Exit from Treatment

IRT Properties of VQIDS-C₅ and VQIDS-SR₅

Assessment of dimensionality of the measures by parallel analysis showed the HRSD₁₇ and IDS-C₃₀ did not satisfy the uni-dimensionality assumption as each scale was determined to have three dimensions and these measures were excluded. All other measures were unidimensional (Tables S1 and S2 of Supplement). Table 2 summarizes the IRT properties of both VQIDS ratings. The “A” value measures the strength of association between the item and the overall amount of depression (larger values indicate a stronger association). The “B” parameters (B1, B2, B3) indicate the level of depression at which there is a 50% probability of being at one level versus higher levels. Thus, the B1 parameter indicates the standardized level of depression at which a participant is as likely to be at “0” versus “1”, “2”, or “3”. The B2 parameter indicates the level at which “0” or “1” is as likely to be endorsed as “2” or “3”. Likewise, the B3 parameter indicates the level at which “0”, “1”, or “2” is as likely as “3”. Higher values of the “B” parameters indicate that levels of the item are endorsed by those with higher levels of depression.

Table 2 VQIDS-C5 and VQIDS-SR5 Item Response Theory Parameters (n=2493)

Figure 1 shows the test information function for all of the measures deemed to be unidimensional. All measures provide the most precise estimates of depression around a theta of one, which corresponds to a level of depression severity of about 1 standard deviation above average. Between 1 and 2 standard deviations above average, QIDS-SR₁₆ had the most precise estimates followed by QIDS-C₁₆, HRSD₆, VQIDS-SR₅, and VQIDS-C₅.

Figure 1 Test information function for all unidimensional measures. Abbreviations: HRSD6, 6-item Hamilton Rating Scale for Depression; QIDS-C16, 16-item Quick Inventory of Depressive Symptomatology – Clinician rated; QIDS-SR16, 16-item Quick Inventory of Depressive Symptomatology – Self-Report; VQIDS-C5, 5-item Very Quick Inventory of Depressive Symptomatology – Clinician rated; VQIDS-SR5, 5-item Very Quick Inventory of Depressive Symptomatology – Self-Report.

Conversion Tables

Table 3 uses the IRT results to provide crosswalks between total scores for the VQIDS-C₅ and the HRSD₆, QIDS-C₁₆, and QIDS-SR₁₆. Table 4 does the same for the VQIDS-SR₅. Table 5 shows how the categories of depressive symptom severity previously established for the QIDS-C₁₆ and QIDS-SR₁₆ are related to the VQIDS-C₅ and VQIDS-SR₅ total scores. For example, a VQIDS-C₅ or a VQIDS-SR₅ score of 6 or more establishes a threshold for at least a moderate depression, as does a 9 or more threshold for severe depression. For the VQIDS-C₅, a 9 reflects a moderate or severe depression, while for the VQIDS-SR₅ a 2 is compatible with no or very mild depression.

Table 3 Conversion Table of VQIDS-C5 to Other Measures (n=2493)

Table 4 Conversion of VQIDS-SR5 to Other Measures (n=2493)

Table 5 Categories of Depressive Symptom Severity

Comparisons for Categories of Benefit

Commonly accepted categories of benefit are defined by remission or symptom changes from baseline. Benefit categories include no meaningful benefit (<35% reduction); partial response (35% to 50% reduction); response without remission (≥50% reduction from baseline but no remission), and remission. These definitions were applied to data compiled at weeks 2, 4, 6 and exit. The remission category was defined as a score of ≤2 on the VQIDS-SR₅ or VQIDS-C₅, a score of ≤5 on the QIDS-C₁₆ or QIDS-SR₁₆, a score of ≤7 on the HRSD₁₇, a score of ≤3 on the HRSD₆, and a score of ≤12 on the IDS-C₃₀. Figure 2A and B show how the QIDS-C₁₆ and VQIDS-C₅, and the QIDS-SR₁₆ and VQIDS-SR₅, respectively, were related in defining the population at various post-baseline measurement occasions. Overall, the longer measure categories were comparable to those formed by the VQIDS categories. Additional information can be found in Supplementary Table 1A-D.

Figure 2 Comparative level of benefit between the QIDS and the VQIDS. (A) QIDS-C16 and VQIDS-C5 level of benefit. (B) QIDS-SR16 and VQIDS-SR5 level of benefit. Abbreviations: QIDS-C16, 16-item Quick Inventory of Depressive Symptomatology – Clinician rated; VQIDS-C5, 5-item Very Quick Inventory of Depressive Symptomatology – Clinician rated; QIDS-SR16, 16-item Quick Inventory of Depressive Symptomatology – Self-Report; VQIDS-SR5, 5-item Very Quick Inventory of Depressive Symptomatology – Self-Report.

Tables 6–8 show the degree of agreement between the QIDS-C₁₆ and VQIDS-C₅, and between the QIDS-SR₁₆ and VQIDS-SR₅, as assessed by kappa statistic for the following categories of benefit: (remission vs no remission, response (with or without remission) vs no response, and benefit vs no benefit) from baseline to week 2, week 4, and week 6, respectively. Table 9 includes the comparison among measures collected at baseline and exit. Overall, there was substantial agreement among measures from baseline to week 2, week 4, and week 6 (ranging from 0.622 to 0.743, 0.669 to 0.749, and 0.709 to 0.780, respectively), and moderate to high agreement among measures collected at baseline and exit (0.599 to 0.866).

Table 6 Strength of Agreement for Remission, Response, and Benefit (Baseline to Week 2) (n=2732)

Table 7 Strength of Agreement for Remission, Response, and Benefit (Baseline to Week 4) (n=2273)

Table 8 Strength of Agreement for Remission, Response, and Benefit (Baseline to Week 6) (n=1984)

Table 9 Strength of Agreement for Remission, Response, and Benefit (Baseline to Exit) (n=2352)

Table 10 shows the comparative sensitivity of the QIDS-C₁₆ and VQIDS-C₅ to change over time as assessed by the percent change from baseline to Weeks 2, 4, 6, and exit. Table 11 does the same for the QIDS-SR₁₆ and the VQIDS-SR₅. Percent change and mean differences among the various measures obtained at baseline and exit can be found in Supplementary Table 2.

Table 10 Percent Change and Mean Differences Among QIDS-C16 - VQIDS-C5 at Different Intervals

Table 11 Percent Change and Mean Differences Among QIDS-SR16 - VQIDS-SR5 at Different Intervals

Discussion

This report, based on a large representative sample of adults with nonpsychotic major depressive disorder treated in primary and psychiatric care settings, found that both the VQIDS-C₅ and VQIDS-SR₅ were unifactorial and had acceptable psychometric properties based on CTT and IRT analyses. Internal consistency (Cronbach’s alpha values) and item-total correlations were acceptable IRT analyses revealed a substantial relationship between each of the 5 items and overall depression severity for the VQIDS-C₅ and even more so for the VQIDS-SR₅. Individual item responses using IRT analysis were also satisfactory.

The IRT analyses also established symptom severity categories (none, mild, moderate, severe, very severe) for each VQIDS based on the accepted symptom severity categories for the QIDS-C₁₆ and QIDS-SR₁₆.⁶ The thresholds used to define the above five severity categories (0–5; 6–10; 11–15; 16–20; 21+) led to virtually identical thresholds for the VQIDS-C₅ and the VQIDS-SR₅. The resulting VQIDS-C₅ categories were 0–2 (none), 3–5 (mild), 6–9 (moderate), 9–12 (severe) and >12 (very severe). The resulting VQIDS-SR₅ categories were 0–2 (none), 2–5 (mild), 6–8 (moderate), 9–12 (severe) and >12 (very severe).

The QIDS-C₁₆ and VQIDS-C₅ identified comparable proportions of patients at exit with the commonly accepted categories of no benefit (19.6% vs 18.2%), partial response (11.2% vs 8.0%), response without remission (19.4% vs 22.9%) and remission (49.9% vs 50.9%), respectively. For the QIDS-SR₁₆ and VQIDS-SR₅, the proportions were also comparable at exit: no benefit (24.1% vs 20.1%), partial response (10.6% vs 7.5%), response without remission (17.5% vs 16.7%), and remission (47.8% vs 55.8%), respectively. Kappa statistics revealed substantial agreement between the VQIDS and the QIDS outcome categories (eg, no meaningful benefit vs some benefit; response vs no response; remission vs no remission, etc).

The study results confirm and extend our initial evaluation of the VQIDS-C₅ and VQIDS-SR₅ in a sample of opportunity of chronically and/or recurrently depressed outpatients treated in primary care or specialty settings.⁵ Both this and our prior report establish that the VQIDS-C₅ and the VQIDS-SR₅ are unifactorial and have acceptable psychometric properties. Further, each VQIDS is as sensitive to change as the corresponding (clinician or self-report) longer QIDS ratings, based on the comparison of outcome categories at 2, 4, and 6 weeks post-baseline.

These results have implications for practice. Since the clinician and self-report versions of the VQIDS were very closely related in this report and in our prior report, it is reasonable to expect that the self-report would be sufficient in most adult depressed outpatients to estimate the overall severity of depressive symptoms as opposed to the potentially more time-consuming clinician-rated version. Since each 5-item VQIDS is as sensitive to change as the longer 16-item (9 domains) QIDS, the VQIDS could be used as a quick check on symptom status, even via smartphone.

The categories of benefit are essential to clinical decision-making and with either VQIDS these categories align nicely with those established for the QIDS. Thus, when clinicians need to decide to what degree the patient has benefited, either the VQIDS self-report or clinician rating is sufficient. In addition, the degree of benefit can be translated into commonly accepted thresholds for improvement using the QIDS. Interestingly, the VQIDS-SR₅ identified a slightly greater proportion of participants that fell under the response, response without remission, and remission categories as early as week 2, which is reassuring in that the self-report is seemingly not more sluggish in detecting change than the clinician rating.

In busy practices, the VQIDS could be used weekly for remote monitoring or telePsychiatry visits, which could shorten the time to make treatment type and dosing decisions, especially if combined with a brief measure of global side effect burden such as the the FIBSER. That is, the VQIDS could facilitate the implementation of MBC. That a self-report could suffice, at least in outpatients, is reassuring because patients can then learn to monitor their own depressions and provide that information in an easily transmitted form whether by paper or digital means. In addition, for those who have greater challenges in self-report, family members can use these tools to assess patient status and communicate with their clinician in a timely fashion.

While the VQIDS (SR or C) were not designed to screen for MDE, their use in this regard deserves study. The scales include both of the essential symptoms (interest and sadness). If they are both scored 0, an MDE cannot be present. Both VQIDS can also assess whether there is a sufficient degree of symptom severity for mild MDE (VQIDS of ≥3) or moderate (VQIDS ≥6) depression (see Tables 3–5 above). The performance of the VQIDS as screening tools deserves study.

There are several study limitations. Both VQIDS ratings were based on items extracted from the QIDS parent rating, which likely inflates the agreement between the VQIDS and QIDS total scores. A comparison of the VQIDS-C₅ and VQIDS-SR₅ with other depression ratings, such as the Montgomery–Åsberg Depression Rating Scale²⁹ or the Beck Depression Inventory,³⁰ at each visit over the course of acute treatment is needed to compare the VQIDS sensitivity to change against longer, more widely accepted depressive symptom ratings beyond the QIDS itself. Further, the generalizability of present results is limited to adult outpatients with non-psychotic major depression. It remains to be determined whether the VQIDS-C₅ and VQIDS-SR₅ display analogous psychometric properties in patients with psychotic or bipolar depression, inpatients, adolescents, and those over 75 years of age. Although the VQIDS scales were unidimensional it remains to be seen if they exhibit longitudinal measurement invariance.³¹ Further, there are no established population norms for either the VQIDS-C₅ or VQIDS-SR₅. A further minor limitation is that the visit-wise QIDS-C₁₆ and QIDS-SR₁₆, and thus the VQIDS derived from them, were collected by clinicians or participants entering their responses on a laptop keyboard.³² Thus, neither a paper and pencil nor a smartphone version has been tested.

Conclusions

In summary, the VQIDS-SR₅ and the VQIDS-C₅ are unifactorial scales with acceptable psychometric properties that provide reliable, clinically informative estimates of overall depressive symptom severity. Each measure is as sensitive to change as the corresponding QIDS-SR₁₆ and QIDS-C₁₆ from which each VQIDS was derived. Either version is suitable as a brief tool to monitor depressive symptoms over time.