How the LBL Depression Test actually works.

1. Instrument selection & naming

The screening questionnaire used in this tool was developed by Kroenke, Spitzer, and Williams in 2001, published in Journal of General Internal Medicine (16:606-613). It is a 9-item self-report instrument deriving directly from the 9 DSM-IV major depressive episode criteria, with each item rated 0-3 on a frequency scale over the past 2 weeks. It is the most widely-used brief depression screener in the world, cited across psychology, sleep medicine, primary care, occupational health, and psychiatry.

Why this instrument over alternatives

1. Brevity. 9 items completed in 2-3 minutes. Beck Depression Inventory has 21 items and is impractical for primary-care screening; Hamilton Depression Rating Scale is clinician-administered.
2. Validation depth. Cronbach's α = 0.89 in original (Kroenke 2001). Sensitivity 88%, specificity 88% at cutoff ≥10 for major depression. Test-retest reliability r = 0.84.
3. Standard cutpoints. 0-4 / 5-9 / 10-14 / 15-19 / 20-27 are the bands Kroenke 2001 published. NHS Talking Therapies (formerly IAPT, Improving Access to Psychological Therapies) uses the instrument as a routine outcome measure, the VA/DoD depression guideline uses it for measurement-based care, and the American Psychiatric Association's DSM-5 Severity Measure for Depression—Adult is adapted from it.
4. Permissive licensing. Pfizer (which funded the original PRIME-MD project) explicitly grants "No permission required to reproduce, translate, display or distribute."
5. Direct DSM mapping. Each of the 9 items corresponds to one of the 9 DSM major depressive episode criteria — symptom-by-symptom transparency.
6. Built-in suicidality item. Item 9 captures self-harm ideation directly — this enables crisis-aware design that screeners without such an item (the HADS, for example) cannot match.

2. The 9 items + functional impairment item

The standard implementation includes the 9 symptom items (scored, sum 0-27) plus a 10th functional-difficulty question that is not added to the score but informs interpretation.

Functional impairment item (not added to score):

If you checked off any problems, how difficult have these problems made it for you to do your work, take care of things at home, or get along with other people?

Response options: Not difficult at all (0) / Somewhat difficult (1) / Very difficult (2) / Extremely difficult (3). This item is captured but not summed. It informs the results copy: a moderate-band score with "extremely difficult" functional impairment is interpreted as warranting more urgent follow-up than a moderate score with "not difficult at all."

Reference period: Past 2 weeks. Response options for items 1-9: 0 = Not at all, 1 = Several days, 2 = More than half the days, 3 = Nearly every day. Total score range: 0-27.

3. Scoring algorithm

The scoring algorithm follows Kroenke 2001 directly. Pseudocode for the implementation:

The implementation is straightforward: integer addition with no weights, no missing-data imputation (the user must answer all 9 items to enable the submit button), and clear band cutoffs.

4. Five severity bands

The 5-band structure comes directly from Kroenke 2001. We do not modify the cutpoints.

5. The cutoff debate: ≥10 vs ≥8

The standard cutoff of ≥10 has been validated repeatedly. Manea, Gilbody & McMillan's 2012 meta-analysis of 18 studies found the instrument performs acceptably at cut-offs from 8 to 11, with similar pooled sensitivity and specificity across that range; 10 sits in that range and is the conventional cut-off.

Some research contexts use ≥8 to maximize sensitivity in screening protocols where missing a case is more costly than over-flagging (e.g., perinatal depression screening). For most clinical and self-screen contexts, ≥10 remains the most-cited and clinically established cutpoint, and is what this tool uses for the moderate-band threshold.

6. Item 9 hard-escalation modal architecture

The item

Item 9 reads: "Thoughts that you would be better off dead, or of hurting yourself in some way." Response scale 0-3 (Not at all / Several days / More than half the days / Nearly every day).

Behavior specification

• Trigger 1: Modal opens immediately when user clicks any non-zero response (1, 2, or 3) on item 9 — before submit.
• Trigger 2: Modal opens on submit if item 9 ≥ 1 and modal has not yet been shown (handles the case where the user clicked rapidly past).
• Re-open: If the user changes their item 9 answer to a different non-zero value, the modal re-opens. This allows resources to be re-displayed if the user reconsiders.
• Dismiss: Modal cannot be dismissed without explicit "I've read this — continue to my results" button click. No X-close icon. No backdrop-click-to-dismiss. Body overflow is locked while modal is open.
• Tier 1 (item 9 = 1): "You said something important." Soft acknowledgment + emergency callout at bottom.
• Tier 2 (item 9 ≥ 2): "What you described is something many people have help available for — and it is treatable." Emergency callout at TOP.
• Resources displayed: 988 (US, call/text), Crisis Text Line HOME→741741, Samaritans UK 116 123, the 9-8-8 Suicide Crisis Helpline in Canada (call/text 9-8-8), findahelpline.com global. Emergency services language: 911 / 999 / 112.
• Non-blocking: User can continue to results after acknowledging.
• Persistent crisis bar: A separate, always-visible crisis bar at the top of every page (tool, methodology, glossary) provides 988 / 741741 / "More resources" links from the moment the page loads. The item-9 modal is the in-tool escalation; the crisis bar is the always-on baseline.

Rationale

Mann et al. 2005's systematic review of suicide prevention strategies found the clearest evidence for restricting access to lethal means and for training physicians to recognize and treat depression. Surfacing crisis resources immediately upon any non-zero response on item 9 — and requiring explicit acknowledgment — is a deliberate friction designed to ensure the user encounters the resources, not to obstruct their use of the tool. The non-blocking design respects user autonomy: anyone can continue to results, but no one inadvertently dismisses the resource panel.

7. Sub-dimension symptom profile

Three dimensions (3/4/1 split)

Item 9 (self-harm ideation) is not included in any sub-dimension grouping. It triggers crisis escalation and is reported as a singleton in the results.

Why these three dimensions specifically

Each dimension is paired with intervention classes that have evidence for depression in general; the pairing itself is LifeByLogic's own, and the cited studies did not test it:

• Cognitive-emotional dominant → CBT for depression, cognitive restructuring, IPT (interpersonal therapy) — Cuijpers et al. 2013 meta-analysis of CBT for depression (115 studies, g ≈ 0.71 against control conditions).
• Somatic dominant → behavioral activation, exercise, sleep stabilization, SSRIs/SNRIs — Cooney et al. 2013 (exercise) and Hollon 2005 (relapse after cognitive therapy or medication).
• Anhedonic-motivational dominant → behavioral activation specifically, novel reward exposure, Brief Behavioral Activation Treatment for Depression (BATD; Lejuez 2001) is a structured 8-15 session protocol with strong RCT support.

Why this is presented as an author choice rather than a validated subscale

Factor analyses of the instrument mostly find a single-factor structure (Cameron 2008; Kocalevent 2013). Some studies, including a large psychiatric sample (Beard 2016), find a 2-factor solution (cognitive-affective + somatic), but no replicated 3-factor structure exists in the literature. The 3-dimension grouping in this tool is interpretive — useful for matching pattern to evidence-based intervention class, but not a validated psychometric subscale.

8. Asymmetric scaling for archetype matching

The three sub-dimensions have asymmetric maximum scores (cognitive-emotional max=9, somatic max=12, anhedonic max=3) because they have asymmetric item counts (3 / 4 / 1). For archetype matching to work — i.e., for any of the three dimensions to be capable of dominating when it is the leading edge — they need to be brought to comparable scale.

The scaling factors

• Cognitive scaled by 1.0 (baseline, max 9)
• Somatic scaled by 0.75 (= 9/12, brings max to 9)
• Anhedonic scaled by 3.0 (= 9/3, brings max to 9)

Why asymmetric scaling is necessary

Without scaling, somatic dominance would be over-easy to achieve (4 items at max gets 12 vs cognitive's max of 9), and anhedonic dominance would be near-impossible to achieve (max 3 vs cognitive's 9). The result would be archetype assignment biased toward "Depleted" (somatic-dominant) and against "Disconnected" (anhedonic-dominant), even when the underlying symptom pattern doesn't warrant it.

Mathematical example

User scores: cognitive=6, somatic=8, anhedonic=2 (total=16).

• Without scaling: somatic 8 > cognitive 6 > anhedonic 2 → "The Depleted"
• With scaling: somatic 8×0.75=6.0; cognitive 6×1.0=6.0; anhedonic 2×3.0=6.0 → tied; first-match-wins logic returns "The Inner Critic" (cognitive)

The scaling puts the three dimensions on the same footing, so the leading dimension, not the item count, decides the archetype; a tie goes to the archetype checked first. This is documented as an author choice because no published precedent for these specific scaling factors exists; they are derived from the asymmetric maximums alone.

9. Four-archetype framework

Four archetypes match symptom profile to evidence-based intervention pathways. They are checked in this order and the first match wins, so a tie for the highest dimension goes to the archetype listed first.

Pathway count per archetype

Each archetype carries 5-7 pathway recommendations, each naming the source it draws on, from meta-analyses and trials to clinical guidelines and books:

• The Steady: 5 pathways (maintenance practices, re-screen, strengthen what's working, read about risk factors, build support map)
• The Inner Critic: 7 pathways (CBT-D, cognitive restructuring, MBCT, IPT, compassion-focused therapy, internet-delivered CBT, defer to clinician at moderate+)
• The Depleted: 7 pathways (behavioral activation, cardiovascular exercise, sleep stabilization, SSRI/SNRI consultation, light therapy, nutritional review, caffeine/alcohol audit)
• The Disconnected: 6 pathways (BATD specifically, activity scheduling with reward monitoring, novel reward exposure, social re-engagement, mindfulness for noticing pleasure, prescriber consultation about other treatments)

Total: 25 evidence-cited pathways across the 4 archetypes.

10. Diagnostic probability per band

Derivation

Values worked out from Kroenke 2001 (sensitivity 88%, specificity 88% at the ≥10 cutoff in n=580 with structured diagnostic interview), assuming that 5 to 9 in 100 people have major depression. At that base rate, about 28 to 42 in 100 people who score 10 or more would have major depression (shown as roughly 1 in 3), and about 1 in 100 of those who score below 10. The tool uses only these figures at the ≥10 cutoff, so it shows one rate below 10 and one at 10 or more rather than a separate rate for each band. They are labeled as approximate and worked out from Kroenke 2001 on the tool page.

Population-dependence caveat

These probability values depend on the base rate of MDE in the user's population. In primary care (where about 5 to 9 in 100 patients have major depression), they are roughly accurate. In psychiatric samples, where the base rate is much higher, the rates would be higher. In low-base-rate populations (e.g., asymptomatic community samples), they would be lower. The values serve as an intuitive interpretation aid, not a precise probabilistic statement applicable to every user.

11. Population norms

The tool compares user score against two reference populations, the German general population (Kocalevent 2013) and US clinic patients without a depressive disorder (Kroenke 2001); the other rows are for reference only:

The general-population norm (Kocalevent 2013) anchors comparison against people not seeking help; the second norm (Kroenke 2001) anchors against the 474 adults in its interview sample, from US primary-care and obstetrics-gynecology clinics, who had no depressive disorder. Both are reported on the tool page with z-scores and percentiles computed via standard normal CDF.

12. Documented author choices summary

Six components of this tool are not directly derivable from published literature and are documented here as transparent author choices:

13. Validation evidence

Internal consistency

Cronbach's α = 0.89 in the original Kroenke 2001 primary-care study (3,000 patients) and 0.86 in its obstetrics-gynecology study (3,000 patients). Kocalevent et al. 2013 found a single factor in a German general-population sample (n=5,018). This is solid internal consistency for a 9-item instrument.

Test-retest reliability

Test-retest correlation r = 0.84 in Kroenke 2001 (about 48 hours between administrations, the second by a mental health professional).

Diagnostic accuracy at ≥10 cutoff

Construct and convergent validity

In Kroenke 2001, functional status fell on all six SF-20 scales as scores rose. In Cameron 2008, the instrument converged more with the depression subscale of the Hospital Anxiety and Depression Scale (HADS) than with its anxiety subscale. It correlates appropriately with anxiety (GAD-7 r ≈ 0.6 — high, reflecting genuine comorbidity, but distinct construct).

Cross-cultural validity

The instrument has been validated in dozens of cross-cultural studies (English, German, Spanish, Mandarin Chinese, Japanese, Arabic, Brazilian Portuguese, French, others). Cultural expression of depression varies, and somatic-vs-cognitive emphasis differs across cultures, but the instrument's basic psychometric properties hold across translations.

14. Functional impairment override

The 10th item (functional difficulty) is captured but not added to the score. It influences the results in one way:

Functional override note

If functional difficulty = "extremely difficult" (value 3) AND total score ≥ 5 (any band above minimal), the results panel surfaces a contextual note:

"You indicated these symptoms have been extremely difficult to live with. That impairment level — regardless of the precise band score — warrants a conversation with a clinician. Difficulty functioning is itself a clinical signal, separate from the symptom-frequency score."

Rationale

Functional impairment is a separate clinical signal from symptom frequency. DSM-5 MDE criteria require "clinically significant distress or functional impairment" — not just symptom presence. A user with mild symptom-frequency but extreme functional difficulty is in a clinically more concerning state than a user with the same symptom score and minimal functional difficulty. The override note surfaces this without modifying the underlying severity-band assignment.

15. Care-aware infrastructure

Depression carries higher absolute mortality risk than anxiety, primarily through suicide. The care-aware infrastructure is consequently more prominent than in the LBL Anxiety Test:

Always-on (independent of any score)

• Persistent crisis bar at the top of every page (tool, methodology, glossary). Visible from the moment the page loads. Provides 988 / 741741 / "More resources" links. Cannot be dismissed.

Score-triggered

• Care-aware top block at score ≥ 10 — surfaces above the results when moderate or higher
• Second care-aware block at score ≥ 10 — sits directly under the severity score, above the rest of the results
• Severe-band clinician prompt at score ≥ 20 (different threshold from the Anxiety Test's ≥ 15 prompt because PHQ has 5 bands not 4)

Item-9 specific (the key escalation path)

• Crisis modal triggered immediately when item 9 ≥ 1. Modal cannot be dismissed without explicit acknowledgment.
• Modal copy escalation at item 9 ≥ 2 (tier-2 emergency callout moves to top)
• Item-9 readout in results panel with red-bordered styling when non-zero

Functional impairment override

• Surfaces at functional = "extremely difficult" AND total ≥ 5, regardless of band

16. Privacy & data handling

Browser-local by default. Specifically:

• The 9 item responses and the functional impairment response are never sent anywhere
• Your total score and its band leave the device only if you choose to save your result to a free LifeGraph+ account
• Your answers and result are never put in localStorage or cookies; if you choose to save your result before signing in, this tab holds it until it is saved or you close the tab
• Closing the tab clears your answers and result from this browser
• The tool uses Google Analytics 4 in a privacy-respecting manner for aggregate metrics only (page views and one completion event) — never individual responses. IP anonymization is enabled.
• Beyond the page view, one GA event is fired, with no PII: tool_complete, which never includes your responses, your score or your band
• If the user copies their share text using the "copy summary" button, item 9 details are deliberately omitted from the shareable text for privacy

17. Limitations

1. Snapshot, not trajectory. The instrument measures last-2-weeks symptoms. Depression naturally fluctuates with life events. A high score during a difficult period doesn't necessarily mean MDE; a low score during a calm period doesn't mean you've never had a depression problem.
2. Self-report dependent. Honest self-report is the foundation. Strong self-criticism, alexithymia, denial, or recall bias can all distort scores in either direction.
3. Not diagnostic. MDE diagnosis requires DSM-5 or ICD-11 clinical interview by a qualified professional. The screen is sensitive (88% per Kroenke 2001) but not diagnostic.
4. Cultural variation. The instrument was developed and first studied in US primary-care and obstetrics-gynecology clinics. Validation in non-Western populations exists but cultural expression of depression varies.
5. Differential diagnosis ignored. A high score may reflect bipolar depression rather than unipolar, persistent depressive disorder rather than MDE, adjustment disorder, or grief. The screen flags "depression symptoms present" not "the source or kind of depression."
6. Sub-dimension scoring is interpretive, not validated. The 3-dimension profile is an author-derived interpretive aid. Use as a conversation starter, not a clinical typology.
7. Archetypes are interpretive frameworks. The 4 archetypes are LBL author-derived. They are not a published clinical typology.
8. Functional impairment item is captured but not summed. The 10th question informs interpretation but does not change the 0-27 score.
9. Validation in adolescents and children is separate. The instrument was studied separately in 442 young people aged 13 to 17 (Richardson 2010), where its best cutoff was 11, not 10. This tool is not validated for users under 18.
10. Documented author choices. Six components are author choices (sub-dimension assignments, asymmetric scaling factors, approximate dx-prob values, the archetype framework and the two item-9 thresholds). All are documented in §12.

18. Frequently asked questions

Why use the 9-item Kroenke 2001 instrument and not something longer like the Beck Depression Inventory?

The 9-item depression screener developed by Kroenke, Spitzer & Williams in 2001 is purpose-built for primary-care screening contexts where time is limited. Its brevity (2-3 minutes), strong validation (Cronbach α=0.89; sensitivity 88%, specificity 88% at cutoff ≥10 per the original validation), and direct mapping to all 9 DSM major depressive episode criteria make it the most widely-used brief depression screener in the world. The Beck Depression Inventory has 21 items; the Hamilton Depression Rating Scale is clinician-administered. Both have their place, but neither is appropriate for a 2-3 minute self-screen.

How were the severity bands chosen?

Directly from Kroenke et al. 2001: 0-4 minimal, 5-9 mild, 10-14 moderate, 15-19 moderately severe, 20-27 severe. NHS Talking Therapies (formerly IAPT, Improving Access to Psychological Therapies) uses the instrument as a routine outcome measure, the VA/DoD depression guideline uses it for measurement-based care, and the American Psychiatric Association's DSM-5 Severity Measure for Depression—Adult is adapted from it. Manea, Gilbody & McMillan's 2012 meta-analysis of 18 studies found the instrument performs acceptably at cut-offs from 8 to 11, with similar pooled sensitivity and specificity across that range; 10 sits in that range and is the conventional cut-off.

Why is the sub-dimension symptom profile considered an author choice?

The original 9-item screener is largely unidimensional in factor analysis (one factor in Cameron 2008 and Kocalevent 2013). Splitting items into cognitive-emotional, physical/somatic, and pleasure-motivation dimensions is an interpretive aid added by LifeByLogic to surface action-relevant patterns. It is presented transparently as an author choice rather than a validated subscale, and the methodology section documents the item assignments and asymmetric scaling factors used.

Why does item 9 trigger an immediate crisis modal that cannot be dismissed casually?

Item 9 captures self-harm ideation. Mann et al. 2005's systematic review of suicide prevention strategies found the clearest evidence for restricting access to lethal means and for training physicians to recognize and treat depression. Surfacing crisis resources immediately upon any non-zero response on item 9 — and requiring explicit acknowledgment to dismiss — is a deliberate friction designed to ensure the user encounters the resources, not to obstruct their use of the tool. The modal is non-blocking; the user can continue to results after acknowledging.

How were the 4 archetypes derived?

The 4 archetypes (The Steady, The Inner Critic, The Depleted, The Disconnected) are LBL author-derived interpretive frameworks that match symptom profile to evidence-based intervention classes. Each archetype maps to interventions with strong meta-analytic or RCT support for depression in general (the pairing with a dimension is LifeByLogic's own) — CBT for cognitive-emotional dominance (Cuijpers 2013), behavioral activation for anhedonic dominance (Dimidjian 2006, Ekers 2014), exercise/sleep stabilization for somatic dominance (Cooney 2013). Archetypes are explicitly named as interpretive frameworks, not a published clinical typology.

Why are the diagnostic probability values labeled 'approximate'?

The values (about 1 in 100 below 10 and roughly 1 in 3 at 10 or more) are worked out from the 88% sensitivity and 88% specificity Kroenke 2001 reports at the ≥10 cutoff in a primary care sample of n=580, assuming that 5 to 9 in 100 people have major depression. They are population-dependent — the actual confirmed-MDE rate in any individual user's case depends on the base rate of major depression in their context. They serve as an intuitive interpretation aid, not a precise probabilistic statement, and the methodology page documents the derivation transparently.

Why include a functional impairment item if it's not added to the score?

The 10th item asking 'how difficult have these problems made it' is part of the original Kroenke 2001 instrument and is captured for the same reason: functional impairment is a separate clinical signal from symptom frequency. A user with moderate symptom-frequency but extreme functional difficulty may need more urgent follow-up than a user with the same symptom score and minimal functional difficulty. This tool surfaces a 'functional override' note in the results when functional difficulty is reported as 'extremely difficult' alongside any symptom score above minimal.

How does this tool handle privacy?

Browser-local by default. The 9 responses and the functional impairment response are never sent anywhere, and your score and its band leave the device only if you choose to save your result. Closing the tab clears your answers and result from this browser. The tool uses Google Analytics 4 in a privacy-respecting manner for aggregate metrics only (page views and one completion event) — never individual responses.

What is this tool not for?

Not a diagnosis. Not a treatment plan. Not a substitute for clinical judgment. Not appropriate for users currently in psychiatric crisis (the crisis modal will surface the relevant resources immediately, but anyone in active crisis should call 988 or local emergency services). Not validated for use under age 18 (the original instrument was validated in adult primary care). Not a longitudinal tracking tool — this is a self-screen at a single moment in time.

How does this differ from the LBL Anxiety Test?

The Anxiety Test uses the 7-item GAD-7 (Spitzer 2006) and produces a 0-21 score with 4 severity bands. The Depression Test uses the 9-item depression screener (Kroenke 2001) and produces a 0-27 score with 5 severity bands. The Depression Test adds an item-9 hard-escalation crisis modal because of the self-harm ideation item, which has no parallel in the GAD-7. Anxiety and depression often occur together (Kroenke 2007); both screens together give a fuller picture than either alone.

19. References & citations

1. Kroenke, K., Spitzer, R. L., & Williams, J. B. W. (2001). The PHQ-9: Validity of a brief depression severity measure. Journal of General Internal Medicine, 16(9), 606–613. doi.org/10.1046/j.1525-1497.2001.016009606.x
2. Manea, L., Gilbody, S., & McMillan, D. (2012). Optimal cut-off score for diagnosing depression with the Patient Health Questionnaire (PHQ-9): a meta-analysis. CMAJ, 184(3), E191–E196. doi.org/10.1503/cmaj.110829
3. Kocalevent, R. D., Hinz, A., & Brähler, E. (2013). Standardization of the depression screener Patient Health Questionnaire (PHQ-9) in the general population. General Hospital Psychiatry, 35(5), 551–555. doi.org/10.1016/j.genhosppsych.2013.04.006
4. Beard, C., Hsu, K. J., Rifkin, L. S., Busch, A. B., & Björgvinsson, T. (2016). Validation of the PHQ-9 in a psychiatric sample. Journal of Affective Disorders, 193, 267–273. doi.org/10.1016/j.jad.2015.12.075
5. Hasin, D. S., Sarvet, A. L., Meyers, J. L., et al. (2018). Epidemiology of adult DSM-5 major depressive disorder and its specifiers in the United States. JAMA Psychiatry, 75(4), 336–346. doi.org/10.1001/jamapsychiatry.2017.4602
6. Cuijpers, P., Berking, M., Andersson, G., Quigley, L., Kleiboer, A., & Dobson, K. S. (2013). A meta-analysis of cognitive-behavioural therapy for adult depression. Canadian Journal of Psychiatry, 58(7), 376–385. doi.org/10.1177/070674371305800702
7. Cooney, G. M., Dwan, K., Greig, C. A., et al. (2013). Exercise for depression. Cochrane Database of Systematic Reviews, (9), CD004366. doi.org/10.1002/14651858.CD004366.pub6
8. Dimidjian, S., Hollon, S. D., Dobson, K. S., et al. (2006). Randomized trial of behavioral activation, cognitive therapy, and antidepressant medication in the acute treatment of adults with major depression. Journal of Consulting and Clinical Psychology, 74(4), 658–670. doi.org/10.1037/0022-006X.74.4.658
9. Ekers, D., Webster, L., Van Straten, A., Cuijpers, P., Richards, D., & Gilbody, S. (2014). Behavioural activation for depression: an update of meta-analysis. PLOS ONE, 9(6), e100100. doi.org/10.1371/journal.pone.0100100
10. Hollon, S. D., DeRubeis, R. J., Shelton, R. C., et al. (2005). Prevention of relapse following cognitive therapy vs medications. Archives of General Psychiatry, 62(4), 417–422. doi.org/10.1001/archpsyc.62.4.417
11. Kuyken, W., Warren, F. C., Taylor, R. S., et al. (2016). Efficacy of mindfulness-based cognitive therapy in prevention of depressive relapse. JAMA Psychiatry, 73(6), 565–574. doi.org/10.1001/jamapsychiatry.2016.0076
12. Mann, J. J., Apter, A., Bertolote, J., et al. (2005). Suicide prevention strategies: a systematic review. JAMA, 294(16), 2064–2074. doi.org/10.1001/jama.294.16.2064
13. American Psychiatric Association. (2013). Diagnostic and Statistical Manual of Mental Disorders (5th ed.). DSM-5. APA Publishing.
14. Lejuez, C. W., Hopko, D. R., & Hopko, S. D. (2001). A brief behavioral activation treatment for depression: Treatment manual. Behavior Modification, 25(2), 255–286. doi.org/10.1177/0145445501252005
15. Kroenke, K., Spitzer, R. L., Williams, J. B. W., Monahan, P. O., & Löwe, B. (2007). Anxiety disorders in primary care: prevalence, impairment, comorbidity, and detection. Annals of Internal Medicine, 146(5), 317–325.
16. Cameron, I. M., Crawford, J. R., Lawton, K., & Reid, I. C. (2008). Psychometric comparison of PHQ-9 and HADS for measuring depression severity in primary care. British Journal of General Practice, 58(546), 32–36.
17. Richardson, L. P., McCauley, E., Grossman, D. C., et al. (2010). Evaluation of the Patient Health Questionnaire-9 Item for detecting major depression among adolescents. Pediatrics, 126(6), 1117–1123.

Read the LBL Depression Test tool itself, or browse the depression screener glossary entry.