The Depression Screening Training in Primary Care

Essentially, the goal of the presentation is to describe the research, which involved implementing depression screening training. The training was aimed at screening older populations for depression in primary care, and it was meant for primary care nurses, nursing professionals and physicians. Thus, the main elements of the project were depression screening and the training that was meant to improve it. The patient population of interest was older people. The rest of the presentation will offer an overview of evidence, an action plan, the findings and, eventually, a conclusion.

Clinical Problem

1. Depression prevalence: 4-13%; 4.7% among the older population (National Institute of Mental Health, 2019, par. 1-5)
2. 50% of primary care visits (Ferenchick, 2019, p. 1)
3. Primary care depression screening: 5% (American Psychiatric Association, 2017, par. 1)
4. Early recognition is important (Conradsson et al., 2013, p. 638)
5. Difficult to diagnose in older adults (Espinoza & Unützer, 2015, par. 22)
6. Healthcare professional training as a problem (Al-Qadhi et al., 2014, pp. 4-5)
7. The reason for choosing this topic is mostly connected to the prevalence of depression. Depending on the population, its prevalence can reach up to 13% (National Institute of Mental Health, 2019, par. 1-5). It is notably lower for older people, amounting to 4.7%, but it remains a significant issue that causes distress, reduces the quality of life and demands an increase in healthcare spending. Overall, the problem of depression is significant enough to dedicate a project to addressing it.
8. Furthermore, the problem of depression screening, especially in primary care, is notable as well. Thus, depending on the study, up to 50% of primary care visits were shown to be associated with depression (Ferenchick, 2019, p. 1), but depression screening only takes place in 5% of visits according to the data for 2017 (American Psychiatric Association, 2017, par. 1). Given that it is important to screen for and recognize depression early (for improved outcomes) important (Conradsson et al., 2013, p. 638), this difference in numbers appears to be an issue. Moreover, specifically for older adults, depression is difficult to diagnose (Espinoza & Unützer, 2015, par. 22). Thus, the clinical problem of depression diagnosing in older adults during primary care visits is worthy of investigation.
9. Finally, it should be pointed out that healthcare professional training to address depression screening is not always sufficient (Al-Qadhi et al., 2014, pp. 4-5). Therefore, all the critical elements of the project are justified by their significance.
10. The widespread nature of depression (National Institute of Mental Health, 2019, par. 1-5)
11. The widespread nature in primary care (American Psychiatric Association, 2017, par. 1)
12. The difficulty of diagnosing in older adults (Espinoza & Unützer, 2015, par. 22)
13. Understudied topic (O’Connor et al., 2016, par. 5-7)
14. Indeed, research clearly shows that the selected mental illness is widespread (National Institute of Mental Health, 2019, par. 1-5), especially in primary care visits (American Psychiatric Association, 2017, par. 1), but its diagnosing is difficult Espinoza & Unützer, 2015, par. 22), which is why the training of healthcare professionals is significant (O’Connor et al., 2016, par. 5-7). Furthermore, research shows that the presented topic is not studied very well, especially as far as depression screening in older adults is concerned (O’Connor et al., 2016, par. 5-7). Overall, research clearly justifies the selected elements and explains why the described clinical problem was chosen.

Significance

Terms

• Depression = major depressive disorder; clinically significant depressed mood, diminished interest and pleasure, fatigue, feelings of worthlessness, etc. (American Psychiatric Association, 2013, p. 161).
• Depression screening: the process of identifying disease (American Psychiatric Association, 2017, par. 1).
• GDS and PHQ-9 (Pfizer Inc., 1999, p. 1; Yesavage et al., 1982, p. 39): common tools for depression screening.
• Older patients: patients aged over 65 (here).

This slide presents the key terms of the presentation. As you can see, the important part here is that older patients were identified as being 65 or older, and the main screening tools of interest were the Geriatric Depression Scale-30 (GDS-30) and Patient Health Questionnaire-9 (PHQ-9). They were chosen for their validity and reliability, as well as their suitability for the task; in particular, GDS has always been meant for older adults, and PHQ is an exceptionally well-studied questionnaire, even though it is not specifically meant for older adults (Pfizer Inc., 1999, p. 1; Yesavage et al., 1982, p. 39). The differences between the two will be discussed in future slides.

PICOT Question

Consider the research question of the project. It is formatted as a PICOT question. As a result, it incorporates the population, which was primary care practitioners, outcome, which was the number of referrals to mental health services, comparisons, which the pre- and post-intervention periods as measured with the help of the two tools, intervention, which was the training, and time, which involved six weeks. Thus, the project took six weeks, included training primary care practitioners to screen older patients for depression with the help of PHQ and GDS, and compared the pre-training and post-training number of referrals to mental health services among the recruited older patients.

Purpose

In line with the question, the purpose of the project consisted of developing and implementing a depression screening program. It was supposed to be meant for primary care professionals who worked with geriatric populations, and it targeted the use of PHQ and GDS. The purpose was based on a literature review, which will be discussed in the next section.

1. MEDLINE, Cochrane, CINAHL, and PubMed
2. Recent, peer-reviewed, relevant (based on an overview of abstracts/full text)

Review of Evidence: Search Method

The flowchart of the evidence review is presented in the slide. The project included 10 articles that were recent, peer-reviewed, and extracted from large medical databases. Only the articles that could be accessed in full were selected; irrelevant papers were excluded.

Review of Evidence: Search Results

In the end, the selected materials included four articles that discussed PHQ and GDS and six articles that covered the varied aspects of depression screening to a different extent, including its significance, processes, and training.

Evidence Synthesis

Depression screening:

1. underdiagnosed, esp. in the elderly (Akincigil & Matthews, 2017; Thabet et al., 2020, p. 1);
2. insufficient/conflicting evidence on benefits (O’Connor et al., 2016; Pfoh et al., 2020; Randle, Spurlock & Kelley, 2019; Rhee, Capistrant, Schommer, Hadsall & Uden, 2017)

Now, the synthesis of all the evidence from those 10 articles will be presented. The synthesis included the topics of depression screening, depression screening training, and depression screening tools. Regarding the first topic, depression screening has been found to be lacking across the globe, including the US. Depression, as a result, remains underdiagnosed, and depending on the population, the issue may be more or less severe. In the case of the older population, the situation is worse than with younger people; depression in older people is especially likely to be underdiagnosed, and diagnosing it in older adults is especially difficult (Akincigil & Matthews, 2017; Thabet et al., 2020, p. 1). Furthermore, it is not clear if depression screening has actual, measurable positive outcomes for patients because the evidence on the topic remains conflicting (O’Connor et al., 2016; Pfoh et al., 2020; Randle, Spurlock & Kelley, 2019; Rhee, Capistrant, Schommer, Hadsall & Uden, 2017). More research is required in that regard, but some evidence does suggest improved patient outcomes that are associated with depression screening in middle-aged adults (O’Connor et al., 2016). For older adults, the data are not conclusive.

Evidence Synthesis (Cont.)

Depression screening training:

1. no research;
2. positive results in primary care practitioners working with adolescents (Fallucco, Seago, Cuffe, Kraemer, & Wysocki, 2015).
3. PHQ-9/GDS: valid, reliable (Conradsson et al., 2013, p. 641; El-Den, Chen, Gan, Wong, & O’Reilly, 2017, p. 509).

Another topic of the literature review was that of depression screening training; that is, training practitioners to screen patients for depression. The search showed that no studies on training practitioners to screen older people have been conducted, but overall, depression screening training is not unlikely to have positive effects, including improved knowledge and confidence, which is proven, for instance, by the study by Fallucco et al. (2015). Admittedly, the study used a sample of practitioners working with adolescents. Still, in the absence of the literature that would be directly relevant, this option has to be used.

Additionally, some research was reviewed to show that PHQ-9 and GDS are valid and reliable with Cronbach’s alpha achieving up to 0.89 (Conradsson et al., 2013, p. 641; El-Den, Chen, Gan, Wong, & O’Reilly, 2017, p. 509). From this perspective, it was important that the two tools differ a lot both in terms of their content and approach to assessment. Thus, PHQ-9 contains 9 questions, and GDS contains 30, which lets the latter be more detailed and the former easier to apply. However, PHQ uses a Likert scale, which is why GDS’s binary options of Yes/No are easier to use for screening and interpreting. Additionally, GDS is meant specifically for older adults. However, PHQ is very well-studied, which is why it was a good choice for the project. Overall, the literature review helped to demonstrate the suitability of the tools that were selected.

Evidence Strengths and Limitations

Strengths:

1. Peer-reviewed
2. Recent
3. High-quality (meta-analyses; quantitative studies) (Polit & Beck, 2017, p. 28)
4. Large samples

Weaknesses:

1. Lack of evidence
2. Generalization/applicability issues
3. The presented evidence has significant strengths. First, all the selected literature was recent and high-quality peer-reviewed articles published in journals. The majority of the literature were quantitative studies, as well as meta-analyses and systematic reviews; according to the evidence pyramid by Polit and Beck (2017), these types of research offer very high-level evidence. It is also noteworthy that all the studies had very sound and detailed methodologies, and the samples were large; most of them were nationally representative.
4. That said, the evidence is still limited mostly because there is a lack of it. For example, the meta-analysis by O’Connor et al. (2016) points out that the data on depression screening in older adults would be too scarce to attempt to make a conclusion about it. Additionally, while the samples are mostly nationally representative, they might not be applicable to regional or other populations. Thus, the presented evidence allows making some suppositions, but there is a limit to them.

Evidence Application

1. Depression screening might have positive outcomes
2. Depression screening training might have positive outcomes
3. GDS-30 and PHQ-9 are suitable tools
4. A need for more data: research justification
5. The application of the evidence suggests that depression screening might have positive outcomes, even though there is little evidence on that topic, especially as applied to older populations, and the same can be said about training practitioners to employ depression screening. Furthermore, the selected tools are suitable for the task, and the research once again shows that the present project is worthwhile. Thus, with the project justified, it is possible to consider the EBP plan.

Population and Setting

1. Primary care center; stakeholders: its providers, patients, and managers
2. Primary care providers: 10
3. Depression screening required in primary care (American Psychiatric Association, 2017, par. 1).
4. Patients: older patients: 15
5. Prevalence/difficulty in diagnosing (Espinoza & Unützer, 2015, par. 22).
6. The setting of the project was a primary care center, and its main stakeholders were the stakeholders of that center. It was planned to involve ten practitioners because the training was supposed to target them, and since the need for research on depression screening in older adults is noteworthy, 15 older patients were supposed to be recruited as well. Both populations, therefore, are justified, and the need to include them is explained by the relevant research (American Psychiatric Association, 2017, par. 1; Espinoza & Unützer, 2015, par. 22).

Conceptual Framework

• Kurt Lewin’s Change Model
• Common for nursing (Ellis & Abbott, 2018, p. 331; Spear, 2016, pp. 59-60)
• Applicable
• Unfreezing: identifying practices, providing information (training)
• Change: application of GDS-30 and PHQ-9
• Refreezing: outside of the bounds of this research

In terms of the framework, the need for change described by Kurt Lewin was employed. The three-stage change model is simple to use, and it has been employed in nursing, which is why it was considered applicable (Ellis & Abbott, 2018, p. 331; Spear, 2016, pp. 59-60). The training, as well as the identification of the existing practices, were the main unfreezing actions, and the change consisted of the application of the new knowledge by the practitioners. Unfortunately, the coronavirus prevented the project from successfully refreezing the situation.

1. Recruitment/ethics
2. Applying PHQ-9
3. Training (importance of screening, depression specifics, the use of the tools)
4. Applying GDS
5. Transferring the data

The following procedures were involved in the project. First, the participants were recruited; the process involved taking into account the ethical aspects of research with human subjects, which is why informed consent was included. After that, the practitioners were asked to administer PHQ-9 to their patients, and the patients were required to fill their copies out. Then, the training of the practitioners took place; they learned about depression and screening, the importance of the latter and the different tools that could be used with a focus on PHQ-9 and GDS-30. After the training, the participants were required to use the GDS. The project took six weeks, and during the final week, all the data were aggregated for analysis.

1. Dependent variable: the number of patients (referred)
2. Tools of data collection: GDS-30 and PHQ-9 (valid, reliable: Conradsson et al., 2013, p. 641-642; El-Den et al., 2017, p. 509)
3. Demographic information on practitioners: short self-developed questionnaire
4. Descriptive statistics (demographics); inferential statistics (the rest), correlations; a=0.05 (Hollander, Wolfe, & Chicken, 2013, p. 7; Polit & Beck, 2017, p. 409).

As can be seen from the PICOT question, the project’s dependent variable was the number of patients that were referred to mental health evaluations. The referrals were based on the results of GDS-30 and PHQ-9, which is why they became the data collection tools; the research that was reviewed during the overview of the evidence suggests that they are both valid and reliable (Conradsson et al., 2013, p. 641-642; El-Den et al., 2017, p. 509). Additionally, a self-developed short questionnaire was included to collect some demographics of the practitioners; it was not tested for validity or reliability, but it only supplied a little information meant to describe the sample, which is why it was considered acceptable. The data analysis involved descriptive and inferential statistics, as well as correlations; the former were used for demographics, and the latter actually helped to respond to the question.

Organizational Factors

• Permission for the research: obtained
• Assistance from the participants (thank you!)
• Coronavirus pandemic: major issue

Regarding additional factors that affected the project, from the perspective of the organization, there were few issues. The permission to carry out the project was obtained easily, and in addition, a lot of support was offered by the participants. In fact, a few of them helped despite the oncoming pandemic and allowed collecting data from them. However, as the lockdown happened, the data collection had to be stopped because no permission from IRB to conduct online data collection had been received. As a result, the coronavirus became the main restricting factor of the project.

Outcomes: Participant Demographics

• Mostly women (89%); mostly nurses (45%)

Experience:

• With both tools: 56%
• With PHQ-9: 22%
• No experience: 22%

The following information about the demographics was collected. First, the practitioners who participated were predominantly women and nurses, and over half of them had some experience with GDS-30 and PHQ-9. However, two of them had no experience with either, and two more people only knew how to use PHQ-9. Thus, while the majority of the people involved were sufficiently experienced, almost half of them needed some training.

As you can see from this graph, in terms of qualifications, the majority of the participants were nurses or nurse practitioners, but a couple of physicians were involved as well. This way, the sample was diversified, and the training was applied to different practitioners. In the end, because of the coronavirus pandemic, only nine practitioners were recruited.

• 9 patients submitted PHQ; 7 submitted GDS (coronavirus)
• Number of referrals increased
• No significant correlations
• Non-parametric test (sign test): no statistically significant difference

Furthermore, only nine patients submitted PHQ-9, and only seven were able to submit GDS-30. Again, the cause of this decrease in the planned sample was the coronavirus. Still, even with a sample this small, some data have been collected, which were analyzed to check the differences in the number of referrals before and after the training, as well as correlations between the PHQ-9 and GDS-30 scores.

The results of the PHQ-9 screening of nine people can be seen in this histogram. It shows the scores of the patient participants, the highest of which was five. As you can see, the majority of the respondents scored zero or one, which implied that they did not need to be referred for mental health evaluation. The patient who scored five was referred as required by the interpretation of the scale.

On the other hand, the histogram of GDS responses shows that more people scored over 5 points while being assessed with the help of this tool, which led to them being referred to a mental health service. The total number of people here is seven, and three of them had the score of six; additionally, more diversity in scores was demonstrated by the rest of the participants, which might be attributed to the fact that GDS has over three times more questions than PHQ, which offers more opportunity for nuance.

Having received the scores from both PHQ-9 and GDS-30, the project could check them for correlations. With the incredibly small sample of seven people, only nonparametric options for any statistical test were available. The results showed that no statistically significant correlations between the two scores could be found, which makes sense given the different numbers of referrals. However, the limitation of the small sample has to be kept in mind while considering these findings; they do not imply that the two tools do not validate each other.

As can be seen from the histograms, before the training, the majority of the patients were not referred; in fact, only one patient out of nine was referred to receive a mental health evaluation, which means that over 88% of them were not referred. However, after the training of the practitioners, they referred more participants (almost half of them or 42.9%), even though by then, only seven patients could submit their data. The data implies that there were some changes in the number of referrals, as well as the percentage of them, although the differences in the sample sizes, as well as the use of different instruments (PHQ and GDS), should be kept in mind.

Finally, the sign test, which is a non-parametric version of the t-test (Polit & Beck, 2017), was used to check for statistically significant differences in the number of patients referred to a mental health evaluation before and after the training. As can be seen from the table, the results were greater than 0.05, which is the universally accepted significance level (Hollander et al., 2013; Polit & Beck, 2017). In turn, it means that there is no statistically significant difference between the referrals before and after training, which is probably the result of the very small sample.

• In line with the existing research: training benefits (Fallucco et al. 2015, p. 327), increased number of diagnosis (Randle et al., 2019, p. 22)
• Some conflicting evidence on the number of diagnoses (Rhee et al., 2017, p. 108)

To summarize, the number of referrals increased as a result of the training, which is in line with the research that implies that the implementation of screening increases the number of referrals. However, the change was not statistically significant, which can also be in line with the evidence that has been reviewed because a study by Rhee et al. (2017) did not show a statistically significant difference in the number of referrals that resulted from implementing depression screening. Basically, the evidence remains rather conflicting, and given the very small sample that was used in this research, it is not possible to make conclusions about the ability of screening to increase the number of referrals. It is also true that no research on the specific topic which was covered by this project exists. As a result, more research is probably required for any conclusive statements.

• Very small sample (the coronavirus)
• No statistically significant effects; cannot produce conclusions
• No correlations: no conclusions about confirmability
• No impact conclusions

Indeed, the project’s sample is a rather significant limitation, which was partially the result of the coronavirus and partially the outcome of the size of the project’s settings. Furthermore, the project did not find any statistically significant results or correlations, which limits its ability to propose conclusions. Additionally, the project did not consider the impacts of the training beyond the number of referrals, which was also measured with the help of different tools. As a result, significant improvements can be made to the methodology in future research.

Practice

• Some providers lack training => training might be required

Research

• more research
• bigger samples
• more variables
• preparing for issues (coronavirus)

Indeed, more research would be needed for a greater understanding of the project’s topic. That additional research should have bigger samples and, possibly, other outcomes variables to figure out what effects the training can have. Also, the experience of this project shows the importance of preparation; the project did not have the permission to employ online means, which is why the sample was even smaller than was originally intended. Future research might use this experience as a learning point and prepare for similar concerns before the project is approved by the IRB. As for practical implications, most findings cannot be used for that, but the research did show that 22% of the providers were not familiar with any of the screening tools, which implies that checking provider readiness to screen and possibly enhancing it via training would be helpful.

• A study dedicated to training practitioners
• PHQ-9 and GDS with older adults in primary care settings
• Justification: underdiagnosed condition in the population; lack of training in the practitioners
• Literature overview: research gap

To summarize, this project was dedicated to training healthcare professionals to screen older adults for depression with the help of PHQ-9 and GDS in primary care. The topic was selected based on a literature review, and it was appropriately justified given the difficulty of diagnosing depression in older adults and the prevalence of depression-related visits in primary care. The possibility of the lack of training in healthcare professionals was also noteworthy; practitioners objectively require training to effectively screen for depression. Additionally, the literature review showed that the project covered a research gap, which further justified it.

• Sample: 9 practitioners; 7 patients (final sample; coronavirus)
• Procedure: Training aimed at using the tools
• Results: Increased number of referrals; no statistical significance/correlations
• Limitations: small sample => difficulty making conclusions
• Practice implications: lack of training
• Need for more research

The final sample of the project was very small; because of the coronavirus, only seven patients managed to submit the GDS. The practitioner sample amounted to nine people; all of them underwent the training meant to help them screen patients and use PHQ and GDS. Some of them also showed a lack of knowledge which further justifies the need for checking provider preparedness and training them to screen for depression.

The results suggested that the number of referrals did increase as a result of the training, although no statistical significance in the change was found. However, with the very small sample, it is next to impossible to make conclusive statements. It can be said that the present study did not prove that training increases referrals in a statistically significant way, but it also did not disprove it. Overall, more research is required for conclusive statements, and given the significance of the chosen topic, more research would be very helpful for primary care practitioners and their patients, especially those of older age.