Running head: SCREENING OLDER ADULTS

Screening Older Adults for Depression in Primary Care
Troy Riutta, RN, BSN
Ann Guthery. PhD, PMHNP-BC
Arizona State University

1

Running head: SCREENING OLDER ADULTS

Screening Older Adults for Depression in Primary Care

2

SCREENING OLDER ADULTS

2
Abstract
Background and Purpose

Depression in older adults is a significant problem that often goes undetected and untreated in
primary care. The U.S. Preventive Services Task Force recommends screening adults for
depression in primary care to increase detection, so it can be adequately managed. Despite this
recommendation, screening rates in primary care are low. The purpose of this project was to
implement a screening intervention and examine the effect of screening on the treatment of
depression in older adults.
Methods
The screening intervention was implemented as an evidence-based project in a small primary
care practice. Consenting adults ≥ 65 years of age were screened with the Patient Health
Questionnaire-9 (PHQ-9). Research indicates the PHQ-9 is valid and reliable for older adults. A
post-screening chart audit was conducted to collect data and analyze the outcome of screening
related to treatment.
Conclusions
A total of 38 participants were screened. Five (13.2%) participants had a positive screening, two
received treatment during the follow up period. The number of participants who were treated
after a positive screening was significant (p= .040).
Implications for Practice
Screening can increase detection and treatment of depression and reduce the associated illness
burden in the older adult population.

Keywords: depression screening, older adults, treatment, primary care

SCREENING OLDER ADULTS

3

Screening older adults for depression in primary care
Since the turn of the century, depression has garnered considerable attention in health
care. Recognition of the negative impact depression has on individuals and the health system
prompted researchers and experts to explore strategies for mitigating the problem. These efforts,
and research highlighting the prevalence of unrecognized depression among adults managed by
general practitioners, led to a focus on depression in the primary care setting.
Problem Statement
Depression is a well-known problem among older adults in the primary care. Despite
awareness of this problem, a significant portion of this population remains undiagnosed. In the
United States, approximately 7 million adults over age 65 are affected by depression (Centers for
Disease Control, 2012). Per the Centers for Medicare and Medicaid Services (CMS) (2011), up
to 25% of older adults with comorbid medical conditions experience depression. Meanwhile, an
estimated 15.1% of Arizona seniors have depression (United Health Foundation, 2016).
Research indicates Primary Care Providers (PCP) detect 40% to 50% of depression cases
among older adults (Park & Unutzer, 2011). Failure to recognize depression in primary care is
significant because older adults tend to receive treatment in this setting. Up to two-thirds of
older adults treated for depression receive the treatment from their primary care provider
(O’Conner, Whitlock, Beil & Gaynes, 2009). Thus, failure to detect is a significant barrier to
treatment.
Undiagnosed and untreated depression associates with significant individual and societal
burden. Personal consequences of depression include suffering, diminished quality of life, and
increased risk of suicide (American Medical Association, 2015). The suicide rate among

SCREENING OLDER ADULTS

4

Arizona residents over 65 was 22.7 per 100,000 residents in 2016 compared to 16 per 100,000
for this age group in the greater U.S. (United Health Foundation, 2016).
Societal implications of depression originate from the burden imposed on the greater
health system. Elderly patients with depression and co-morbid chronic conditions have
significantly higher health care costs compared to those without depression due to increased
utilization including frequent hospitalization, emergency department visits, and doctor visits
(Centers for Disease Control, 2012). Depression also complicates the treatment and outcome of
co-morbid medical conditions. The Centers for Disease Control (CDC) estimates the economic
cost of depression from lost productivity, health expenditures, and suicide exceeded 200 billion
dollars in 2010 (CDC, 2016).
In the last several years, awareness of the depression burden led to a shift in public
health policy toward a focus on recognizing depression in primary care. In 2002, the U.S.
Preventive Service Task Force (USPSTF) recommended depression screening for adults in
primary care (USPSTF, 2002). The USPSTF recommendation influenced the 2011 Medicare
decision to pay for annual depression screenings in primary care (CMS, 2011). Clinical practice
guidelines from the Agency for Healthcare Research and Quality (AHRQ) recommend routine
depression screening for all adults in primary care (AHRQ, 2016).
Purpose
As the major entry point for health care services, primary care providers (PCP) are in a
unique position to detect and treat depression in the older adult population. The purpose of this
paper is to discuss the background and significance of the problem, the strategy used to search
the literature, synthesize the evidence, and discuss an evidence-based project that implemented
the evidence into practice.

SCREENING OLDER ADULTS

5

Background and Significance
Screening older adults with a standardized age-appropriate instrument can help
PCPs detect depression in this population. In a systematic review, Watson and Pignone
(2003) examined the accuracy of depression screening instruments in the older adult
primary care population and identified several instruments with good sensitivity and
specificity including the Geriatric Depression Scale (GDS), Center for Epidemiological
Studies Depression Scale, and the SelfCARE(D). This review influenced the original
USPSTF recommendation.
O’Conner, Whitlock, Beil, and Gaynes (2009) expanded upon this research by
conducting a systematic review to aid the USPSTF in updating its original 2002
screening recommendation. The authors concluded screening helps identify depression
but screening alone does not improve outcomes (O’Conner, Whitlock, Beil & Gaynes,
2009). Instead, screening as part of a greater approach that includes additional support
and treatment is beneficial (O’Conner, Whitlock, Beil & Gaynes, 2009). These findings
influenced the revised USPSTF recommendations in 2009.
In a recent systematic review, O’Conner et al. (2016) examined evidence
regarding benefits and harms of depression screening in the adult population to provide a
comprehensive updated review for the USPSTF. The rationale for routine screening is to
identify undiagnosed depression and reduce the amount of time between depression
onset, and initiation of treatment (O’Conner et al., 2016). The authors concluded
depression screening programs increase the likelihood of remission or reduction of
depression symptoms in response to treatment (O’Conner et al., 2016).

SCREENING OLDER ADULTS

6

An increase in diagnosis and treatment from screening is significant considering the
prolonged time it typically takes people to receive treatment after onset of symptoms. According
to Pence, O’Donnell, and Gaynes (2012), the median time from symptom onset to depression
treatment is eight years. Meanwhile, O’Conner et al (2016) report the average onset to treatment
time is four years.
Evidence also indicates screening and subsequent treatment of depression can reduce
suicide. A systematic review of suicide prevention programs reveals depression screening and
awareness programs correlate with reduced risk of suicide in older adults (Lapierre et al., 2011).
Furthermore, screening and treatment interventions can reduce non-suicide related mortality.
Gallo et al. (2013) found that older adults who were screened, and received depression treatment
interventions, had a lower mortality risk compared to those receiving usual care; mortality risk
among those in the depression treatment group was similar to people without depression.
According to the USPSTF (2015), both medication and psychotherapy can increase rates of
remission from depression.
A common theme in the literature regarding depression screening is the relation between
screening and treatment. Specifically, the goal of screening in primary care is to detect
depression so it can be treated. Depression is treatable in up to 80% of cases (CDC, 2012).
Despite evidence supporting depression screening in primary care, expert
recommendation, and initiatives from public health authorities, it appears screening is not yet
standard practice. Data from the National Ambulatory Medical Care Survey from 2005 to 2010
reveals an average annual screening rate of 1.8% in primary care (McGoey, Huang & Palmes,
2013). Meanwhile, Maimone and Mahartta (2015) analyzed a random sample of 500 patients
and identified a slightly higher screening rate of 14.6% (Maimone and Mahartta, 2015). While

SCREENING OLDER ADULTS

7

estimates of screening rates may vary, the low rates indicate screening practices are
under-utilized in primary care.
The results of this inquiry, indicating screening helps detect depression and can
result in subsequent treatment, leads to the following clinically relevant PICOT question:
“In older adult primary care patients, how does depression screening, compared to no
screening, affect treatment over a three-month period?”
Search Strategy
An extensive search of research databases was conducted to identify evidence
pertinent to the clinical question. This entailed an exhaustive search of three databases:
PsycINFO, PubMED, and CINAHL. Each database search entailed use of keywords and
Boolean connectors. The keywords used are depression screening, older adults or
elderly, primary care or general practice, and treatment or referral. Title and abstracts
of results from each database were reviewed and relevant articles saved for further
appraisal. A total of 28 articles were identified from the search. Each study was
critically appraised according to the inclusion and exclusion criteria. Ten final studies
relevant to the clinical question were included in an evaluation table (Appendix A).
Critical Appraisal and Synthesis of Evidence
Studies contained in this review are generally high quality studies with varying
levels of evidence. Six studies represent level III evidence or higher, and the remainder
are level IV evidence. All studies are quantitative and include depression screening as a
component, either alone, or as an integral part of a broader intervention. There is
significant heterogeneity among interventions across studies ranging from screening only,
to screening with feedback, and intensive support. Outcome variables for the level I

SCREENING OLDER ADULTS

8

through III studies include depression response or symptom reduction, remission, and mortality.
Meanwhile, lower level studies examined screening rates and correlations between screening,
diagnosis, treatment, and outcomes.
Overall, the studies reported an adequate amount of demographic data, employed
appropriate methods, and had a clearly-defined research purpose. All but three studies had a
greater number of female than male participants. Four studies focused exclusively on older
adults while the remainder contained the general adult population including older adults.
Evidence from the six level I through III studies reveal varying degrees of support for
depression screening. The effect of screening is weakest when used alone and gets stronger
when combined with additional support including treatment. This finding was consistent for
outcome variables across all the higher-level studies. Five of the six studies report screening
interventions improved depression symptoms, four report increased rates of remission, in three
there was a reduction in suicidal ideation, and two report a reduction in depression related
mortality risk. Studies included in the systematic reviews were well designed RCT’s and cohort
studies with adequate follow-up time, randomization, blinding and concealment, acceptable
attrition rates, and homogenous sample characteristics. However, some studies included in the
systematic reviews were underpowered with wide confidence intervals and failed to find even
large improvements in depressive symptoms significant. In contrast, findings from adequatelypowered studies were significant and more precise with narrow confidence intervals.
Results from the level IV studies were consistent for similar outcomes. Screening rates
were low across all studies. Three studies showed higher rates of diagnosis from screening and
in two of these studies, screening was associated with initiating treatment. These studies were
retrospective, had large sample sizes, utilized data from electronic health records, and employed

SCREENING OLDER ADULTS

9

appropriate measures including regression and variable analysis to limit confounding
factors and bias. The PHQ-9 and HAM-d were the most commonly used screening
instruments across studies.
Conclusion from the Evidence
Screening older adults for depression in primary care increases detection of the
illness and can lead to diagnosis and subsequent treatment. When screening results in
diagnosis and treatment, it improves outcomes by reducing symptoms, increasing rates of
remission, and lowering the risk of depression related mortality.
Project Aims
The purpose of this evidence-based project is twofold: implement the USPSTF
screening recommendation to screen older adults for depression in primary care and
examine the effect of screening on treatment.
EBP Model for Guiding Project
The Stetler Model of Evidence-Based Practice provides a guide for implementing
the depression screening intervention. The model consists of distinct phases that guide a
change in practice: identifying a need for change, assessing the evidence, making
decisions about the evidence, incorporating the evidence into a plan for use and
implementing the plan, and evaluating whether the goals related to implementation were
met (Melnyk & Fineout-Overholt, 2015). The model can be used to implement change
by an individual, group, or entire organization (Stetler, 2001). The flexibility of the
Stetler Model makes it a good fit for implementing a depression screening in a small
primary care practice.

SCREENING OLDER ADULTS

10
Methods

Human Subjects Protection
Approval for this project was granted by the Arizona State University Institutional
Review Board and the project site. Participation in the project was voluntary and no personally
identifying data was collected.
Setting and Participants
The project was implemented at a small urban primary care practice in Arizona. Patients
were eligible to participate if they were ≥ 65 years of age, English speaking, able to give consent,
and had an appointment on the day of screening. Patients under age 65, unable to give consent,
and non-English speaking were excluded.
Design and Procedures
The project contained two phases: a depression screening intervention and a follow up
chart audit. During the screening phase, eligible patients were given a cover letter explaining the
purpose of the project and time was taken to answer questions. Consenting patients were
screened with the PHQ-9 and results communicated to both the patient and the PCP. The PCP
documented the screening result in the electronic medical record (EMR).
Three months after the screening intervention, in February 2018, the chart audit was
completed. To facilitate the audit, the EMR was utilized to query all patients who had an
appointment on the dates of screening. A documented screening result from the date of
screening connected the patient to the project.
Measures
The PHQ-9 was derived from the Primary Care Evaluation of Mental Disorders Patient
Health Questionnaire and designed for use in primary care settings (Kroenke, Spitzer &

SCREENING OLDER ADULTS

11

Williams, 2001). Items on the PHQ-9 are based on the diagnostic criteria for depression
from the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV); it can be used
as an initial screening tool and to assess the ongoing severity of depression (Kroenke,
Spitzer & Williams, 2001). The instrument demonstrates validity and reliability when
used on older adults with a sensitivity of 88% and specificity of 80% (Phelan et al, 2010).
A score ≥10 is considered a positive screening (Kroenke, Spitzer & Williams, 2001).
Thus, for project purposes, this was the score used to determine a positive screening.
Data Collection and Analysis
Data collection occurred during the chart audit. Relevant data was transferred
onto a data collection spreadsheet in de-identified form to protect the privacy of
participants. Specific data included demographic information such as age and gender,
screening result, diagnosis status before and after screening, and treatment status before
and after screening.
Descriptive statistics were used to describe the sample and outcome variables. A
two-tailed test was used to analyze the data and the critical value set at p<.05.

For

evaluation purposes, treatment was defined as either initiating a new treatment or
changing an existing treatment after the screening. SPSS was used to store and manage
the data (cite SPSS).
Results
The sample (N=38) consisted of 23 females (60.5%) and 15 males (39.5%)
participants. Mean age was 79.7 (SD=7.8), the minimum age was 66 and maximum age
in the 90 range. The maximum score on the PHQ-9 was 16 with a mean score of 3.4
(SD=4.4) among all participants. Five participants (13.2%) screened positive for

SCREENING OLDER ADULTS

12

depression and 32 (86.8%) were negative. Twelve (31.6%) participants had an existing
diagnosis of depression. One participant was lost to follow up because there was no documented
screening result, and thus, no way to connect the participant to the project.
Of the five who screened positive, two received treatment after the screening. Due to the
small sample size, Fisher’s exact test was used to test the data for statistical significance. At an
alpha of 0.05, the proportion of participants who received treatment after screening is significant
(p=.040). Participants were more likely to get treatment after a positive screening (OR= 21.3,
95% CI [1.47-310]).
Discussion
The purpose of this project was to implement an evidence-based depression screening
and examine how screening effects treatment. Results of the analysis indicate there was a small,
but statistically significant, increase in treatment after the screening intervention. This finding
indicates screening can be an effective way to increase diagnosis and treatment of depression in
older adults.
Several themes emerged from the data analysis. First, the prevalence of depression
among participants is similar to the greater population. An estimated 15% of Arizona seniors
suffer from depression (United Health Foundation, 2016). In comparison, 13.8% of participants
screened positive for depression. Next, there was a greater number of female than male
participants. This is a phenomenon found in most of the studies reviewed from the literature.
Another theme to emerge from the data is that screening serves multiple purposes in the
management of depression. Kroenke (2012) reports the PHQ-9 is a multipurpose instrument
ideal for screening for undetected depression, gauging the severity of depression, and monitoring
the effectiveness of treatment. This concept is evident when examining the screening results.

SCREENING OLDER ADULTS

13

Four of the 5 participants with a positive screening had existing diagnosis and treatment.
For these individuals, screening positive suggested their treatment was not effective.
Meanwhile, 8 participants with a negative screening had an existing diagnosis and
treatment. Some of these participants scored zero on the PHQ-9 which indicates
remission of symptoms. In these cases, the negative screening reflects adequate
treatment.
Implications
The findings of this project support the use of screening in primary care to detect
depression and increase the number of older adults who receive treatment. These
findings add to the existing body of evidence supporting the USPSTF screening
recommendation. Providers can screen patients with an age appropriate instrument like
the PHQ-9 to detect depression, gauge the severity of known depression, and evaluate the
effectiveness of treatment.
It is important to note that this study was conducted over a short period of time
and focused solely on the relationship between screening and treatment. As a result, it
was not possible to evaluate the adequacy of treatment and follow up. This is important
because there is evidence that depression is often inadequately treated in primary care.
According to Park and Unutzer (2011), most older adults treated for depression receive
the treatment from their PCP but only around 20% are adequately treated. Consequently,
future studies should examine the adequacy of treatment over time. In addition, larger
studies are needed to better estimate the effect of screening.

SCREENING OLDER ADULTS

14

Limitations
A major limitation of this study is the small sample size. This makes the estimated effect
of screening less precise and is evident in the wide confidence interval. A small sample size also
makes the findings less generalizable to the greater population. Therefore, the results should be
interpreted within the context of these limitations.
Conclusion
Despite the limitations of this project, the results suggest screening has an important role
in addressing depression. Screening older adults for depression can increase detection of the
illness and lead to diagnosis and treatment. Adequate treatment can reduce the burden
depression imposes on individuals by improving quality of life and minimizing its adverse effect
on co-morbid medical conditions. In addition, a higher number of older adults treated for
depression may reduce the impact on society associated with increased health costs from higher
utilization of health services.

SCREENING OLDER ADULTS

15
References

Agency for Healthcare Research and Quality (2016). Adult depression in primary care.
Retrieved from https://www.guideline.gov/summaries/summary/50406
Akincigil, A. & Matthews, E. (2017). National rates and patterns of depression screening in
primary care: Results from 2012 and 2013. Psychiatric Services. doi:
10.1176/appi.ps.201600096
American Medical Association (2015). Preventive care and screening: Screening for clinical
depression and follow up plan- national quality strategy domain: community/ population
health. Retrieved from http://www.aana.com/resources2/qualityreimbursement/Documents/2016_PQRS_Measure_134_11_17_2015.pdf
Burton, C., Simpson, C. & Anderson, N. (2012). Diagnosis and treatment of depression
following routine screening in patients with coronary heart disease or diabetes: a database
cohort study. Psychological Medicine, 43, 529-537. doi: 10.1017/s0033291712001481
Centers for Disease Control and Prevention (2012). CDC promotes public health approach to
address depression among older adults. Retrieved from
https://www.cdc.gov/aging/pdf/cib_mental_health.pdf
Centers for Disease Control and Prevention (2016). Depression. Retrieved from
https://www.cdc.gov/mentalhealth/basics/mental-illness/depression.htm
Centers for Medicare and Medicaid (2011). Decision memo for screening for depression in
adults. Retrieved from https://www.cms.gov/medicare-coverage-database/details/ncadecision-memo.aspx?NCAId=251

SCREENING OLDER ADULTS

16

Centers for Medicare and Medicaid (2016). Comprehensive primary care initiative. Retrieved
from https://innovation.cms.gov/files/x/cpci-ecqm-manual2016.pdf
Gallo, J., Morales, K., Bogner, H., Raue, P., Zee, J., Bruce, M. & Reynolds, C. (2013). Long
term effect of depression care management on mortality in older adults: Follow up of cluster
randomized clinical trial in primary care. British Medical Journal. doi: 10.1136/bmj.f2570
Kroenke, K. (2012). Enhancing the clinical utility of depression screening. Canadian Medical
Association Journal, 184 (3), 281-282. doi: 10.1503/cmaj.112004
Kroenke, K., Spitzer, R. & Williams, J. (2001). The PHQ-9: Validity of a brief depression
severity measure. Journal of General Internal Medicine, 16, 606-613. Retrieved from
https://www.ncbi.nlm.nih.gov/pubmed/11556941
Lapierre, S., Erlangson, A., Margda, W., Diego, D., Oyama, H., Scocco, P., . . . Quinnett, P.
(2011). A systematic review of elderly suicide prevention programs. Crisis, 32, 88-98.
doi:10.1027/0227-5910/a000076
Maimone, R. & Mahartta, A. (2015). The rate of depression screening at a federally qualified
community health center. Health Services Research and Managerial Epidemiology, 1-4.
doi: 10.1177/2333392815613057
McGoey, S., Huang, K. & Palmes, G. (2013). Low depression screening rates in U.S.
ambulatory care. Psychiatric Services, 64, 1068. doi: 10.1176/appi.ps.201300132
Melnyk, B. & Fine-Overholt, E. (2015). Evidence-based practice in nursing & healthcare: A
guide to best practice. New York, NY: Wolters Kluwer Health.
Mojtabai, R (2011). Does depression screening have an effect on the diagnosis and treatment of
mood disorders in general medical settings? An instrumental variable analysis of the

SCREENING OLDER ADULTS

17

national ambulatory medical care survey. Medical Care Research and Review, 68, 462-489.
doi: 10.1177/1077558710388290
O’Conner, E., Whitlock, E., Beil, T. & Gaynes, B. (2009). Screening for depression in adult
patients in primary care settings: A systematic evidence review. Annals of Internal
Medicine, 151, 793-803. doi: 10.7326/0003-4819-151-11-200912010-00007
O’Conner, E., Rossom, R., Henninger, M., Groom, H., Burda, B., Henderson, J., . . . Whitlock,
E. (2016). Screening for depression in adults: An updated systematic evidence review for the
U.S. preventive services task force. Agency for Healthcare Research and Quality. Retrieved
from https://www.ncbi.nlm.nih.gov/pubmed/26937538
Oyama, H. & Sakashita, T. (2016). Long-term effects of a screening intervention for depression
on suicide rates among Japanese community-dwelling older adults. The American Journal of
Geriatric Psychiatry, 24(4), 287-296. doi 10.1016/j.jagp.2015.10.008
Park, M. & Unützer, J. (2011). Geriatric depression in primary care. The Psychiatric clinics of
North America, 34(2). doi10.1016/j.psc.2011.02.009
Pence, B., O’Donnell, J. & Gaynes, B. (2012). The depression treatment cascade in primary
care: A public health perspective. Current Psychiatry Report, 14, 328-335. doi:
10.1007/s11920-012-0274-y
Pfoh, E., Mojtabai, R., Bailey, J., Weiner, J. & Dy, S. (2015). Conformance to depression
process measures of medicare part b beneficiaries in primary practice settings. Journal of the
American Geriatrics Society, 63, 1338-1345. doi: 10.1111/jgs.13483

SCREENING OLDER ADULTS

18

Phelan, E., Williams, B., Meeker, K., Bonn, K., Frederick, J., LoGerfo, J. & Snowden, M.
(2010). A study of the diagnostic accuracy of the PHQ-9 in primary care elderly. BMC
Family Practice, 11, 63. doi: 10.1186/1471-2296-11-63
Pignone, M., Gaynes, B., Rushton, J., Mulrow, C., Orleans, T., Whitener, L., . . .Lohr, K. (2002).
Screening for depression: Systematic evidence review. Annals of Internal Medicine, 136,
765-776. Retrieved from https://www.ahrq.gov/downloads/pub/prevent/pdfser/depser.pdf
Stetler, C. (2001). Updating the settler model of research utilization to facilitate evidence-based
practice. Nursing Outlook, 49, 272-279. doi: :10.1067/mno.2001.120517
United Health Foundation (2016). Americas health rankings: 2016 senior report. Retrieved
from http://www.americashealthrankings.org/explore/2016-seniorreport/measure/depression_sr/state/AZ
U.S. Preventive Services Task Force (2002). Screening for depression: Recommendations and
rationale. American Family Physician, 15, 647-650. Retrieved from
http://www.aafp.org/afp/2002/0815/p647.html
U.S. Preventive Services Task Force (2015). Final recommendation statement-depression in
adults: screening. Retrieved from
https://www.uspreventiveservicestaskforce.org/Page/Document/RecommendationStatementF
inal/depression-in-adults-screening1
Watson, L. & Pignone, M. (2003). Screening accuracy for late-life depression in primary care: A
systematic review. The Journal of Family Practice, 52, 956-964. Retrieved from
https://www.ncbi.nlm.nih.gov/pubmed/14653982

UNDIAGNOSED DEPRESSION

19
Appendix A

Table 1
Evaluation Table
Citation

Theory/
Conceptual
Framework

Design/ Method

Sample/ Setting

Major Variables
& Definitions

Measurement/
Instrumentation

Data Analysis
(stats used)

Findings/
Results

Akincigil &
Matthews (2017).
National rates and
patterns of
depression
screening in
primary care:
Results from 2012
and 2013.

InferredDiffusions of
Innovations
Framework

Cross-sectional
secondary
analysis

n= 33,653 patientphysician
encounters
F>M

IV1- gender
IV2- EHR use
IV3-participation
meaningful use or
federal value
program
IV4- quality
reimbursement

NR

Stata statistical
software,
Pearson Chisquare tests and
Multi-variate
logistic regression
CI=95%

DS rate= 4.2%
47% of screened
got new dx of
MDD

Funded by AHRQ
No conflicts of
interest or bias
noted
USA

Purpose:
examine rates
and patterns of
DS among PCP
visits and identify
associations with
initiatives such as
EHR adoption
through
meaningful use
or quality
initiatives.

IC:
Adults without
existing
depression; PC
visit
EC:
Under age 18;
prior depression
dx

DV-DS

EHR use- 5% DS
(AOR 1.81, p=
.001)
GenderM: 3.8%
: 4.4% (AOR
1.36, p= .110)
Quality
reimbursementNS (AOR 1.09,
p=.087)

Level/Quality of
Evidence;
Decision for
practice/
application to
practice
Level IV
Strengths:
Large sample
size, general
description of
screening
practices,
appropriate
methods
Weaknesses:
Assumed
screening
occurred if new
dx of MDD,
retrospective

Key: AHRQ- Agency for Healthcare Research and Quality; AOR- adjusted odds ratio; ARR- absolute risk reduction; BDI- Beck Depression Inventory; CGControl Group; CI- confidence interval; CIDI- Composite International Diagnosis Interview DS- depression screening; DV-dependent variable; dxdiagnosis; EC- exclusion criteria; EHR- electronic health record; f/u- follow-up; GDS- Geriatric Depression Scale; HAM-d- Hamilton Depression Scale; ICinclusion criteria; IV- independent variable; IG- intervention group; IRR- incident rate ratio; MA- meta-analysis; MDD- major depressive disorder; MHmental health; MMSE- mini mental status exam; N-number of studies; n- number of participants; NAMCS- National Ambulatory Medical Care Survey;
NNT- number needed to treat; OR- odds ratio; PC- primary care; PHQ-9- Patient Health Questionnaire-9; PCP-primary care provider; RI- relative incidence;
RS- random sampling; RR- relative risk; RRR- relative risk reduction; SR-systematic review; tx- treatment; SRDS- Zung Self-Rated Depression Scale;
USPSTF- U.S. Preventive Services Task Force

UNDIAGNOSED DEPRESSION

20

Citation

Theory/
Conceptual
Framework

Design/ Method

Sample/ Setting

Major Variables
& Definitions

Measurement/
Instrumentation

Data Analysis
(stats used)

Findings/
Results

Gallo et al. (2013)

Inferred-Health
Prevention
framework

RCT-long-term
f/u of
PROSPECT
study /
quantitative

n= 1226

CES-D
HAM-D

Age stratified RS

SAS 9.1 and Stata
12.0 Cox
proportional
hazards
regression;
Kaplan & Meier
method

Hazard ratio for
MDD in CG 1.90
(95% CI: 1.572.31)

Purpose:
Assess whether
the increased
mortality risk
among patients
with MDD can
be reduced to the
risk of those
without MDD

IV1-MDD in IG
IV2-No MDD in
IG
IV3-MDD in CG
DV-mortality risk
(risk of death of
patients with
MDD in IG vs
risk in those with
MDD in CG)

Funding: NIMH
Bias/Conflicts:
none
USA

Setting: 20 PC
practices in 3
cities

IC: age ≥ 60;
MMSE score >17;
English speaking;
CES-D score > 20

CES-D tested in
older adults:
sensitivity 83%,
specificity 78%
Unknown if
HAM-D tested in
older adults

CI= 95%

Hazard ratio for
MDD in IG 1.09
(95% CI: 0.831.44)
Hazard ratio for
MDD in IG
compared to
MDD in CG 0.76
(95% CI: 0.571.00)

Conclusion:
Explanatory
significance;
screening rates
are low. Quality
reimbursement
not a predictor of
screening
Level/Quality of
Evidence;
Decision for
practice/
application to
practice
Level II
Strengths:
Large sample
size, long f/u
period, RCT
Weaknesses:
Interventions used
in IG may not be
feasible or cost
effective in
general settings
Conclusions:
Findings indicate
treating MDD in

Key: AHRQ- Agency for Healthcare Research and Quality; AOR- adjusted odds ratio; ARR- absolute risk reduction; BDI- Beck Depression Inventory; CGControl Group; CI- confidence interval; CIDI- Composite International Diagnosis Interview DS- depression screening; DV-dependent variable; dxdiagnosis; EC- exclusion criteria; EHR- electronic health record; f/u- follow-up; GDS- Geriatric Depression Scale; HAM-d- Hamilton Depression Scale; ICinclusion criteria; IV- independent variable; IG- intervention group; IRR- incident rate ratio; MA- meta-analysis; MDD- major depressive disorder; MHmental health; MMSE- mini mental status exam; N-number of studies; n- number of participants; NAMCS- National Ambulatory Medical Care Survey;
NNT- number needed to treat; OR- odds ratio; PC- primary care; PHQ-9- Patient Health Questionnaire-9; PCP-primary care provider; RI- relative incidence;
RS- random sampling; RR- relative risk; RRR- relative risk reduction; SR-systematic review; tx- treatment; SRDS- Zung Self-Rated Depression Scale;
USPSTF- U.S. Preventive Services Task Force

UNDIAGNOSED DEPRESSION

21

Citation

Theory/
Conceptual
Framework

Design/ Method

Sample/ Setting

Major Variables
& Definitions

Measurement/
Instrumentation

Data Analysis
(stats used)

Findings/
Results

Burton (2012).
Diagnosis and
treatment of
depression
following routine
screening in
patients with
coronary heart
disease or
diabetes: a
database cohort
study.

Quality and
Outcomes
Framework

RCS /
quantitative

n=67,358
Data from General
Practice
Administration
System in UK
F>M

IV-DS

PHQ-9

Self-control, caseseries method

DV1- dx within 4
weeks
DV2-tx within 4
weeks

PHQ-9 tested in
older adults:
sensitivity 88%,
specificity 80%

DV1:
RI 3.03 (95% CI,
2.44-3.78)

Funding:
National Health
Service of
Scotland

Purpose:
Evaluated patient
records to
examine relation
between DS and
dx/tx in the
month following
DS in patients
with chronic
illness

older adults
reduces risk of
mortality/morbidit
y. Aligns with
other studies that
treating MDD has
substantial
benefits.
Level/Quality of
Evidence;
Decision for
practice/
application to
practice
Level IV

Setting:
237 PC practices

Strengths:
Large sample
size, case-control
design reduces
confounding
factors related to
differences
between subjects

IC:
Dx of coronary
heart disease or
diabetes and DS
during study
period

Weaknesses:
Lower level of
evidence,
retrospective
study, did not
include those with

Relative
incidence
95% CI

DV2:
RI 1.78 (95% CI,
1.54-2.05)

Key: AHRQ- Agency for Healthcare Research and Quality; AOR- adjusted odds ratio; ARR- absolute risk reduction; BDI- Beck Depression Inventory; CGControl Group; CI- confidence interval; CIDI- Composite International Diagnosis Interview DS- depression screening; DV-dependent variable; dxdiagnosis; EC- exclusion criteria; EHR- electronic health record; f/u- follow-up; GDS- Geriatric Depression Scale; HAM-d- Hamilton Depression Scale; ICinclusion criteria; IV- independent variable; IG- intervention group; IRR- incident rate ratio; MA- meta-analysis; MDD- major depressive disorder; MHmental health; MMSE- mini mental status exam; N-number of studies; n- number of participants; NAMCS- National Ambulatory Medical Care Survey;
NNT- number needed to treat; OR- odds ratio; PC- primary care; PHQ-9- Patient Health Questionnaire-9; PCP-primary care provider; RI- relative incidence;
RS- random sampling; RR- relative risk; RRR- relative risk reduction; SR-systematic review; tx- treatment; SRDS- Zung Self-Rated Depression Scale;
USPSTF- U.S. Preventive Services Task Force

UNDIAGNOSED DEPRESSION

22

Bias/ Conflicts of
interest: none
identified

DS, dx, and tx on
same day

EC:
DS, dx, and tx on
same day

United Kingdom

Citation

Theory/
Conceptual
Framework

Design/ Method

Sample/ Setting

Major Variables
& Definitions

Measurement/
Instrumentation

Data Analysis
(stats used)

Findings/
Results

Lapierre et al.
(2011).
Systematic review
of elderly suicide
prevention
programs.

Inferred -Health
Prevention
Framework

SR / quantitative

N=19 studies
n= 31,505

IV1- DS
IV2- tx
IV3- education
interventions
IV3- social
isolation
reduction
DV1-suicidal
ideation
DV2-depression
symptoms

Specific
instruments NR

Qualitative
synthesis

DV1:
3/4 studies
showed
significant
reduction in SI (p
< 0.05)
1 study showed
NS reduction in
SI

Funding: not
disclosed
Bias/Conflicts:
none noted

Purpose:
examine results
of interventions
for suicidal
elderly persons to
identify
successful
strategies
Method:
Searched 5

Setting:
Multi-national
IC:
Studies with
empirical
evaluation of a
suicide
intervention or
prevention

DV2:
3/4 studies
showed

Conclusions:
Findings indicate
there was higher
rate of dx and tx
after screening.
Screening is
effective in
people with
chronic illness
Level/Quality of
Evidence;
Decision for
practice/
application to
practice
Level I
Strengths:
Large sample
size, SR, focus on
elderly
Weaknesses:
No MA due to
heterogeneity of
interventions in

Key: AHRQ- Agency for Healthcare Research and Quality; AOR- adjusted odds ratio; ARR- absolute risk reduction; BDI- Beck Depression Inventory; CGControl Group; CI- confidence interval; CIDI- Composite International Diagnosis Interview DS- depression screening; DV-dependent variable; dxdiagnosis; EC- exclusion criteria; EHR- electronic health record; f/u- follow-up; GDS- Geriatric Depression Scale; HAM-d- Hamilton Depression Scale; ICinclusion criteria; IV- independent variable; IG- intervention group; IRR- incident rate ratio; MA- meta-analysis; MDD- major depressive disorder; MHmental health; MMSE- mini mental status exam; N-number of studies; n- number of participants; NAMCS- National Ambulatory Medical Care Survey;
NNT- number needed to treat; OR- odds ratio; PC- primary care; PHQ-9- Patient Health Questionnaire-9; PCP-primary care provider; RI- relative incidence;
RS- random sampling; RR- relative risk; RRR- relative risk reduction; SR-systematic review; tx- treatment; SRDS- Zung Self-Rated Depression Scale;
USPSTF- U.S. Preventive Services Task Force

UNDIAGNOSED DEPRESSION
Country:
International
network of
researchers

23
databases from
1966-2009

program; age ≥60;
peer reviewed

significant
reduction in
symptoms;
1 study showed
no difference
between IG and
CG

Citation

Theory/
Conceptual
Framework

Design/ Method

Sample/ Setting

Major Variables
& Definitions

Measurement/
Instrumentation

Data Analysis
(stats used)

Findings/
Results

Mojtabai (2010).
Does depression
screening have an
effect on the
diagnosis and
treatment of mood
disorders in

Practice Variation
Framework

Observational
study /
quantitative

n=73,712
F>M
RS collected from
NAMCS

IV-DS
DV1- mood
disorder dx
DV2- tx with
antidepressant

NR

Stata 11

DS occurred in
1.8% of sample

Purpose:
examine effects
of DS on dx and

IC: Non psychiatric visits

Wald tests, Rho
co-efficients,
Probit analyses,
adjusted F tests

DV1:
positive
association
between DS and

studies, some
interventions may
not be feasible
Conclusion:
Findings indicate
prevention
programs with DS
as a component
can reduce suicide
related events.
Some
interventions
likely not
generalizable to
every day
practice.
Level/Quality of
Evidence;
Decision for
practice/
application to
practice
Level IV
Strengths:
Large sample
size, detailed
explanation of
methods,

Key: AHRQ- Agency for Healthcare Research and Quality; AOR- adjusted odds ratio; ARR- absolute risk reduction; BDI- Beck Depression Inventory; CGControl Group; CI- confidence interval; CIDI- Composite International Diagnosis Interview DS- depression screening; DV-dependent variable; dxdiagnosis; EC- exclusion criteria; EHR- electronic health record; f/u- follow-up; GDS- Geriatric Depression Scale; HAM-d- Hamilton Depression Scale; ICinclusion criteria; IV- independent variable; IG- intervention group; IRR- incident rate ratio; MA- meta-analysis; MDD- major depressive disorder; MHmental health; MMSE- mini mental status exam; N-number of studies; n- number of participants; NAMCS- National Ambulatory Medical Care Survey;
NNT- number needed to treat; OR- odds ratio; PC- primary care; PHQ-9- Patient Health Questionnaire-9; PCP-primary care provider; RI- relative incidence;
RS- random sampling; RR- relative risk; RRR- relative risk reduction; SR-systematic review; tx- treatment; SRDS- Zung Self-Rated Depression Scale;
USPSTF- U.S. Preventive Services Task Force

UNDIAGNOSED DEPRESSION
general medical
settings? An
instrumental
variable analysis
of the NAMCS.

24
tx decisions in
patients with
MDD in usual
practice

from 2005-2007

95% CI

Funding/Bias/Con
flicts: none
identified

dx (B=0.690,
SE=0.087, p<
0.001)
Probability of Dx
in screened 0.039
vs 0.007 in nonscreened
DV2:
positive
association
between DS and
tx (B=0.545,
SE=0.274, p= <
0.001)
Probability of tx
0.139 in screened
vs 0.052 in nonscreened

USA

Citation

Theory/
Conceptual
Framework

Design/ Method

Sample/ Setting

Major Variables
& Definitions

Measurement/
Instrumentation

Data Analysis
(stats used)

Findings/
Results

O’Conner et al.
(2016).
Screening for
depression in
adults: An

Population Health
framework

SR / quantitative

N=71
n=3,814
F>M

IV1- DS
IV2- feedback
IV3- tx
interventions

PHQ-9; HAM-D;
Geriatric
Depression Scale;
Beck Depression
Inventory; CES-d

Stata 13.1

DV1 / DV2:
DS programs
increased
remission or tx
response 20%-

Purpose:
Systematically
review and

IC:

DerSimonian and
Laird pooled
estimate;

description of
conceptual
framework
Weaknesses:
NAMCS dataunable to
independently
verify dx,
screening
instruments not
identified
Conclusion:
DS helps identify
MDD and
increases
probability of tx
Level/Quality of
Evidence;
Decision for
practice/
application to
practice
Level I
Strengths:
Sound methods,

Key: AHRQ- Agency for Healthcare Research and Quality; AOR- adjusted odds ratio; ARR- absolute risk reduction; BDI- Beck Depression Inventory; CGControl Group; CI- confidence interval; CIDI- Composite International Diagnosis Interview DS- depression screening; DV-dependent variable; dxdiagnosis; EC- exclusion criteria; EHR- electronic health record; f/u- follow-up; GDS- Geriatric Depression Scale; HAM-d- Hamilton Depression Scale; ICinclusion criteria; IV- independent variable; IG- intervention group; IRR- incident rate ratio; MA- meta-analysis; MDD- major depressive disorder; MHmental health; MMSE- mini mental status exam; N-number of studies; n- number of participants; NAMCS- National Ambulatory Medical Care Survey;
NNT- number needed to treat; OR- odds ratio; PC- primary care; PHQ-9- Patient Health Questionnaire-9; PCP-primary care provider; RI- relative incidence;
RS- random sampling; RR- relative risk; RRR- relative risk reduction; SR-systematic review; tx- treatment; SRDS- Zung Self-Rated Depression Scale;
USPSTF- U.S. Preventive Services Task Force

UNDIAGNOSED DEPRESSION
updated
systematic
evidence review
for the USPSTF.
Funding: AHRQ
Bias/conflicts:
none identified
USA

25
update benefits
and harms of DS
in adults to aid
the USPSTF in
updating its
recommendation
Methods:
Searched 4
databases from
2009-2015;
websites of
government
agencies,
professional
organizations for
grey literature;
reviewed journal
tables of contents

Studies targeting
DS; Brief
standardized
instrument
designed to
identify persons
with depression;
PC settings;
RCTs, CCTs; SR
EC:
Comparative
effectiveness
studies; study
populations with
pre-existing
depression;
intervention is
second line tx;
studies with f/u
period < 6 weeks

DV1- depression
remission
DV2-symptom
reduction
DV3- health
outcomes
Health outcomes:
improved
outcomes defined
as decreased
symptoms;
decreased suicide
deaths, attempts
or ideation;
improved
functioning;
improved quality
of life; improved
health status

Tested for use in
older adults:
PHQ-9:
sensitivity 88%,
specificity 80%
GDS: sensitivity
75%, specificity
70%, CES-d:
sensitivity 83%,
specificity 78%,
BDI: sensitivity
88%, specificity
81.7%
Unknown if tested
in older adults:
HAM-D

Egger’s test;
contingency
tables; Knapp
Hartung
modification;
Forest plots

80%
7/9 studies
showed
improvement in
symptoms or
remission but
only 2 significant:
RR 1.19 (95% CI,
1.06-1.34)
RR 1.71 (95% CI,
1.13-2.57)
NS studies range
from RR 1.13
(95% CI, 0.462.79) to RR1.79
(955 CI, 0.943.41)

large number of
studies, thorough
explanation of
methods, PC
setting, adequate
f/u time among
studies
Weaknesses:
Many studies
underpowered to
detect even large
differences in
treatment effect,
significant
heterogeneity
among study
interventions
prevents metaanalysis, greater
number of
females
Conclusion:
DS programs are
effective in
improving
symptoms and
remission when

Key: AHRQ- Agency for Healthcare Research and Quality; AOR- adjusted odds ratio; ARR- absolute risk reduction; BDI- Beck Depression Inventory; CGControl Group; CI- confidence interval; CIDI- Composite International Diagnosis Interview DS- depression screening; DV-dependent variable; dxdiagnosis; EC- exclusion criteria; EHR- electronic health record; f/u- follow-up; GDS- Geriatric Depression Scale; HAM-d- Hamilton Depression Scale; ICinclusion criteria; IV- independent variable; IG- intervention group; IRR- incident rate ratio; MA- meta-analysis; MDD- major depressive disorder; MHmental health; MMSE- mini mental status exam; N-number of studies; n- number of participants; NAMCS- National Ambulatory Medical Care Survey;
NNT- number needed to treat; OR- odds ratio; PC- primary care; PHQ-9- Patient Health Questionnaire-9; PCP-primary care provider; RI- relative incidence;
RS- random sampling; RR- relative risk; RRR- relative risk reduction; SR-systematic review; tx- treatment; SRDS- Zung Self-Rated Depression Scale;
USPSTF- U.S. Preventive Services Task Force

UNDIAGNOSED DEPRESSION

26
identified MDD
leads to tx

Citation

Theory/
Conceptual
Framework

Design/ Method

Sample/ Setting

Major Variables
& Definitions

Measurement/
Instrumentation

Data Analysis
(stats used)

Findings/
Results

O’Conner et al.
(2009).
Screening for
depression in
adult patients in
primary care
settings: A
systematic
review.

Population Health
Framework

SR / quantitative

N=33
n=12,432

IV1-DS
IV2- DS with
feedback and/or
support
interventions
IV3-tx
interventions
DV1-depressive
symptoms &
remission
DV2-health
status/outcome
DV3-depression
dx

PHQ-9, HAM-D;
CES-d

SAS 8.2

DV1:

Poisson
distribution;
qualitative
synthesis

IV1- 1 study on
effects of DS on
MDD symptoms:
remission
achieved in 48%
screened vs 27%
of non-screened
(p < 0.05).

Funding: AHRQ
Bias/conflicts:
none identified
USA

Purpose:
SR for USPSTF
about benefits
and harms of
adult DS in PC
Methods:
Searched 5
databases from
1998-2007

IC:
Study focus on
DS; outcomes
identified; SR;
RCT; large cohort
studies with
minimum 1000
participants; PC
setting
EC:
Inpatient setting;
intervention not
appropriate for
PC; focus on
children; nonEnglish language;

Definitions:
Health
status/outcomes
defined as
improvement in

Tested for use in
older adults:
PHQ-9:
sensitivity 88%,
specificity 80%,
CES-d: sensitivity
83%, specificity
78%
Unknown if tested
in older adults:
HAM-D

95% CI

IV25/7 studies show
significant
improvement in
symptoms and
remission (p<
0.05);
2/7 show NS
improvement in

Level/Quality of
Evidence;
Decision for
practice/
application to
practice
Level I
Strengths:
Large number of
studies, examined
tx in older adults;
PC setting
Weaknesses:
Heterogeneity
among studies
prevents metaanalysis, mixed
quality of studies
Conclusion:
Findings indicate
screening
programs are
beneficial when

Key: AHRQ- Agency for Healthcare Research and Quality; AOR- adjusted odds ratio; ARR- absolute risk reduction; BDI- Beck Depression Inventory; CGControl Group; CI- confidence interval; CIDI- Composite International Diagnosis Interview DS- depression screening; DV-dependent variable; dxdiagnosis; EC- exclusion criteria; EHR- electronic health record; f/u- follow-up; GDS- Geriatric Depression Scale; HAM-d- Hamilton Depression Scale; ICinclusion criteria; IV- independent variable; IG- intervention group; IRR- incident rate ratio; MA- meta-analysis; MDD- major depressive disorder; MHmental health; MMSE- mini mental status exam; N-number of studies; n- number of participants; NAMCS- National Ambulatory Medical Care Survey;
NNT- number needed to treat; OR- odds ratio; PC- primary care; PHQ-9- Patient Health Questionnaire-9; PCP-primary care provider; RI- relative incidence;
RS- random sampling; RR- relative risk; RRR- relative risk reduction; SR-systematic review; tx- treatment; SRDS- Zung Self-Rated Depression Scale;
USPSTF- U.S. Preventive Services Task Force

UNDIAGNOSED DEPRESSION

27
outcomes not
disclosed; f/u
period < 6 weeks

comorbid illness,
reduction in
physical
symptoms,
reduction in SI

symptoms or
remission

Suicidal
behaviors: acts or
attempts to selfharm

IV3suicidal behaviors
reduced in older
adults with tx, OR
0.06 (95% CI,
0.01-0.58); SI
reduced in older
adult OR 0.39
(95% CI, 0.180.78)
Findings/
Results

DV2:

Citation

Theory/
Conceptual
Framework

Design/ Method

Sample/ Setting

Major Variables
& Definitions

Measurement/
Instrumentation

Data Analysis
(stats used)

Oyama (2016)
Long-term effects
of a screening
intervention for
depression on
suicide rates
among Japanese
communitydwelling older

InferredPopulation
Prevention
framework

CCS /
quantitative

n= 4,918
F>M
Age ≥ 65

IV1- DS
IV2- DS &
education
IV3- usual care

SRDS; GDS;
CIDI

SPSS

Purpose:
Evaluate longterm impact of
routine
depression
screening on

Geographic cohort
sampling
Setting:
Community

Validated in older
adults:

DV1-suicide rate
Definitions:

DS leads to Dx
and tx; tx reduces
SI in older adults

Zung: sensitivity
58%-76%,
specificity 82%-

Mixed-effects
binomial
regression
models; repeatedmeasures linear
models; IRR

48% decrease in
suicide rate in IG
vs no change in
CG
IRR in IG 0.52
(95% CI, 0.330.83, p=0.008)

Level/Quality of
Evidence;
Decision for
practice/
application to
practice
Level III
Strengths:
Large sample
size, long f/u
period
Weaknesses:
Japanese study-

Key: AHRQ- Agency for Healthcare Research and Quality; AOR- adjusted odds ratio; ARR- absolute risk reduction; BDI- Beck Depression Inventory; CGControl Group; CI- confidence interval; CIDI- Composite International Diagnosis Interview DS- depression screening; DV-dependent variable; dxdiagnosis; EC- exclusion criteria; EHR- electronic health record; f/u- follow-up; GDS- Geriatric Depression Scale; HAM-d- Hamilton Depression Scale; ICinclusion criteria; IV- independent variable; IG- intervention group; IRR- incident rate ratio; MA- meta-analysis; MDD- major depressive disorder; MHmental health; MMSE- mini mental status exam; N-number of studies; n- number of participants; NAMCS- National Ambulatory Medical Care Survey;
NNT- number needed to treat; OR- odds ratio; PC- primary care; PHQ-9- Patient Health Questionnaire-9; PCP-primary care provider; RI- relative incidence;
RS- random sampling; RR- relative risk; RRR- relative risk reduction; SR-systematic review; tx- treatment; SRDS- Zung Self-Rated Depression Scale;
USPSTF- U.S. Preventive Services Task Force

UNDIAGNOSED DEPRESSION
adults

28
suicide rates in
older adults

setting in Japan

Bias/Conflicts:
none identified

Usual care: usual
periodic health
check-ups for
general
population

Funding:
Japan Ministry of
Health

86%
GDS: sensitivity
75%, specificity
70%

IRR in CG 0.93
(95% CI, 0.691.26)
Ratio of IRR
between CG and
IG 1.83 (95% CI,
1.08-3.09, p=
0.026)

Unknown if
validated in older
adults:
CIDI

may reflect
cultural factors
not present in
U.S.
Conclusion:
DS programs can
reduce SI and
suicide rate

Japan
Citation

Theory/
Conceptual
Framework

Design/ Method

Sample/ Setting

Major Variables
& Definitions

Measurement/
Instrumentation

Data Analysis
(stats used)

Findings/
Results

Pfoh et al. (2015).
Conformance to
depression
process measures
of medicare part b
beneficiaries in
primary care
settings.

Inferred- Quality
Performance
Framework

Cross-sectional
study using EHR
data / quantitative

N= 34 clinics
n= 5000
F>M

PHQ-9

Stata 12

DS rate 17%

Purpose:
evaluate
conformance to
DS, management,
and outcome
quality indicators
and identify
characteristics
associated with
conformance

Quota RS: half
had annual well
visit and half
routine visit

IV1- annual well
visit
IV2- routine visit
DV-DS
DV2reassessment
within 3 months
after new dx of
MDD
DV3- depression
response (> 50%
reduction in
symptoms
measured by

Tested in older
adults: sensitivity
88%, specificity
80%

Descriptive
analyses;
Multivariate
logistic regression

Odds of DS for
general visit:
AOR 1.16 (95%
CI, 0.86-1.55)
Odds of DS for
well visit: AOR
0.41 (95% CI,
0.30-0.56)
Odds of
depression
response: AOR
3.93 (95% CI,

Funding: AHRQ
Bias/ Conflicts:
none identified

IC: selected
through RS
EC:
DS within 4

Level/Quality of
Evidence;
Decision for
practice/
application to
practice
Level IV
Strengths:
Large sample
size, clear
findings, PC
setting, no bias
noted
Weaknesses:
Lower level
evidence,

Key: AHRQ- Agency for Healthcare Research and Quality; AOR- adjusted odds ratio; ARR- absolute risk reduction; BDI- Beck Depression Inventory; CGControl Group; CI- confidence interval; CIDI- Composite International Diagnosis Interview DS- depression screening; DV-dependent variable; dxdiagnosis; EC- exclusion criteria; EHR- electronic health record; f/u- follow-up; GDS- Geriatric Depression Scale; HAM-d- Hamilton Depression Scale; ICinclusion criteria; IV- independent variable; IG- intervention group; IRR- incident rate ratio; MA- meta-analysis; MDD- major depressive disorder; MHmental health; MMSE- mini mental status exam; N-number of studies; n- number of participants; NAMCS- National Ambulatory Medical Care Survey;
NNT- number needed to treat; OR- odds ratio; PC- primary care; PHQ-9- Patient Health Questionnaire-9; PCP-primary care provider; RI- relative incidence;
RS- random sampling; RR- relative risk; RRR- relative risk reduction; SR-systematic review; tx- treatment; SRDS- Zung Self-Rated Depression Scale;
USPSTF- U.S. Preventive Services Task Force

UNDIAGNOSED DEPRESSION
USA

29
to these
measures.

months;
depression
remission within 1
year

PHQ-9)
DV4- remission
within 12 months
(PHQ-9 score <
5)

1.46-10.57, p <
0.05)
Odds of
reassessment
within 3 months:
0.71 (95% CI,
0.17-3.02)

retrospective data,
use of EHR
records, no
identification of
factors that
contribute to
increased odds of
response
Conclusion:

Citation

Theory/
Conceptual
Framework

Design/ Method

Sample/ Setting

Major Variables
& Definitions

Measurement/Inst
rumentation

Data Analysis
(stats used)

Findings/
Results

Findings indicate
rate of DS is low.
Does show that
people in study
with DS had
greater odds of
response.
Level/Quality of
Evidence;
Decision for
practice/
application to
practice

Key: AHRQ- Agency for Healthcare Research and Quality; AOR- adjusted odds ratio; ARR- absolute risk reduction; BDI- Beck Depression Inventory; CGControl Group; CI- confidence interval; CIDI- Composite International Diagnosis Interview DS- depression screening; DV-dependent variable; dxdiagnosis; EC- exclusion criteria; EHR- electronic health record; f/u- follow-up; GDS- Geriatric Depression Scale; HAM-d- Hamilton Depression Scale; ICinclusion criteria; IV- independent variable; IG- intervention group; IRR- incident rate ratio; MA- meta-analysis; MDD- major depressive disorder; MHmental health; MMSE- mini mental status exam; N-number of studies; n- number of participants; NAMCS- National Ambulatory Medical Care Survey;
NNT- number needed to treat; OR- odds ratio; PC- primary care; PHQ-9- Patient Health Questionnaire-9; PCP-primary care provider; RI- relative incidence;
RS- random sampling; RR- relative risk; RRR- relative risk reduction; SR-systematic review; tx- treatment; SRDS- Zung Self-Rated Depression Scale;
USPSTF- U.S. Preventive Services Task Force

UNDIAGNOSED DEPRESSION
Pignone et al.
(2002)
Screening for
depression:
Systematic
evidence review.
Funding: AHRQ
Bias/Conflicts:
none identified
USA

Population Health
Framework

30
SR / quantitative
Purpose:
Examine
effectiveness of
DS in PC settings
Methods:
Searched 2
databases from
1966-1999;
Reviewed
bibliographies

N= 70
n= 15,263
F>M

IV1-DS
IV2-DS with
feedback

IC:
English language;
RCT; SR/MA; PC
setting; original
research

DV1-dx
DV2-tx
DV3-outcomes

EC:
Non-PC setting;
focus on children;
non-English; no
original data

Outcome defined
as decreased
MDD symptoms,
improved quality
of life, and
reduction in
morbidity/mortalit
y

HAM-d; BDI

Stata 6.0

BDI tested in
older adults:
sensitivity 88%,
specificity 81.7%

DerSimonian and
Laird randomeffects model

Unknown if
HAM-d tested in
older adults

Risk of persistent
MDD after DS:
RR 0.87 (95% CI,
0.79-0.95), RRR
13%; ARR 9%;
NNT achieve
remission at 6
months= 11
DV1:
4/12 studies
showed
significant
increase in dx (p
< 0.05) in IG with
DS vs CG with no
DS;
2/12 studies NS
(p > 0.05);
6/12 studies did
not report
significance data;
Rate of dx from
DS increased
10%-47.5%
DV2:
4/10 studies had
higher rates of tx
in IG vs CG (p <
0.05);

Level I
Strengths:
Large number of
studies, SR, clear
methodology,
clearly stated
research
questions,
detailed
explanation of
included studies
strengths/
weaknesses
Weaknesses:
Level of
significance NR
in some studies.
Heterogeneity of
studies prevented
meta-analysis,
some studies had
missing data
Conclusion:
Findings support
DS in PC. DS
increases
incidence of dx
and tx.

Key: AHRQ- Agency for Healthcare Research and Quality; AOR- adjusted odds ratio; ARR- absolute risk reduction; BDI- Beck Depression Inventory; CGControl Group; CI- confidence interval; CIDI- Composite International Diagnosis Interview DS- depression screening; DV-dependent variable; dxdiagnosis; EC- exclusion criteria; EHR- electronic health record; f/u- follow-up; GDS- Geriatric Depression Scale; HAM-d- Hamilton Depression Scale; ICinclusion criteria; IV- independent variable; IG- intervention group; IRR- incident rate ratio; MA- meta-analysis; MDD- major depressive disorder; MHmental health; MMSE- mini mental status exam; N-number of studies; n- number of participants; NAMCS- National Ambulatory Medical Care Survey;
NNT- number needed to treat; OR- odds ratio; PC- primary care; PHQ-9- Patient Health Questionnaire-9; PCP-primary care provider; RI- relative incidence;
RS- random sampling; RR- relative risk; RRR- relative risk reduction; SR-systematic review; tx- treatment; SRDS- Zung Self-Rated Depression Scale;
USPSTF- U.S. Preventive Services Task Force

UNDIAGNOSED DEPRESSION

31
3/10 had higher tx
in IG vs CG but
NS;
2/10 had higher tx
rates in IG but
significance NR;
1/10 studies had
higher tx rates in
CG than IG
DV3:
In 5/9 studies
significant
improvement in
outcomes (p<
0.05);
In 1/9 studies
improvement NS;
In 3/9 studies,
outcomes
improved but
significance NR

Key: AHRQ- Agency for Healthcare Research and Quality; AOR- adjusted odds ratio; ARR- absolute risk reduction; BDI- Beck Depression Inventory; CGControl Group; CI- confidence interval; CIDI- Composite International Diagnosis Interview DS- depression screening; DV-dependent variable; dxdiagnosis; EC- exclusion criteria; EHR- electronic health record; f/u- follow-up; GDS- Geriatric Depression Scale; HAM-d- Hamilton Depression Scale; ICinclusion criteria; IV- independent variable; IG- intervention group; IRR- incident rate ratio; MA- meta-analysis; MDD- major depressive disorder; MHmental health; MMSE- mini mental status exam; N-number of studies; n- number of participants; NAMCS- National Ambulatory Medical Care Survey;
NNT- number needed to treat; OR- odds ratio; PC- primary care; PHQ-9- Patient Health Questionnaire-9; PCP-primary care provider; RI- relative incidence;
RS- random sampling; RR- relative risk; RRR- relative risk reduction; SR-systematic review; tx- treatment; SRDS- Zung Self-Rated Depression Scale;
USPSTF- U.S. Preventive Services Task Force