Depression is an increasingly widespread issue in the United States (US). It is not treated in primary care, but it is screened for with the help of diverse tools, which can then lead to a mental health referral. Said tools include the Geriatric Depression Scale (GDS) (Yesavage et al., 1982), as well as the Patient Health Questionnaire-9 (PHQ-9) (Pfizer Inc., 1999). In this report, a project dedicated to the training for the use of different tools is going to be described. The report involves an analysis of the existing evidence, an explanation of the specifics and significance of the current inquiry, and an overview of the methodology and outcomes of the project.
Depression is a major healthcare concern in the US and globally (American Psychiatric Association, 2017; Cohen & Edmondson, 2015; Farrer et al., 2016; Fekadu, Shibeshi, & Engindawork, 2017; Ferenchick, 2019). Depression-related visits often (up to 50% of cases) occur in primary care settings (Ferenchick, 2019), but the majority of primary care patients (over 95%) are not screened for depression, which may be partially the cause of low recognition and treatment of depression (American Psychiatric Association, 2017; Park & Unützer, 2011). At the same time, early recognition and treatment of depression are critical for health outcomes in patients (Conradsson et al., 2013). As a result, the problem of the lack of screening in primary care is rather acute.
It is also important that additional difficulties in diagnosis may be present for particular populations. For example, the prevalence of depression among older adults is significantly lower than it is among younger people, with 13.1% of the former population experiencing a major depressive episode in the past year, and only 4.7% of the latter population doing the same (National Institute of Mental Health, 2019). However, older patients experience lower chances of accurate diagnosis because the symptoms of depression, for example, changes in information processing or eating habits, overlap with the natural age-related changes (Espinoza & Unützer, 2015; Zaccagnini & White, 2014). In other words, it is difficult to diagnose depression in older adults, which further highlights the importance of screening for this population.
However, other causes of depression mismanagement at the screening stage include managerial failures (Moriarty, Gilbody, McMillan, & Manea, 2015), as well as the lack of training in healthcare professionals (Al-Qadhi, Rahman, Ferwana, &Abdulmajeed, 2014). Given that the difficulty in depression management that becomes prominent at the old age cannot be altered, the possibility of equipping primary care practitioners with the knowledge and other tools that can facilitate undertaking the challenge is crucial. Thus, the clinical problem of this project is the underscreening for depression in primary care settings as attributable to the lack of training in practitioners.
With around eight million cases in the US each year (Ferenchick, 2019), with up to 13% of the population affected by the disorder over their lifetime (American Psychiatric Association, 2017), depression is a significant, widespread concern. According to the American Psychiatric Association (2017), only about 5% of adults who visit primary care services are screened for depression, which leaves about half the cases unrecognized. Older people are an especially challenging population in terms of depression screening (Espinoza & Unützer, 2015; Zaccagnini & White, 2014). Thus, the problem clearly affects a large portion of the population of the US, with patients with depression and their families being the primary stakeholders.
Review of Evidence
In this section, the PICOT question is going to be developed and contextualized with the help of a review of the existing evidence.
Among practitioners in primary care settings (P), is there any difference in the number of geriatric patients referred to mental health practitioners for evaluation for depression (O) as measured by the Patient Health Questionnaire-(PHQ-9) screening tool and Geriatric Depression Scale (GDS) (C) as a result of training aimed at teaching practitioners to use both tools (I) over the course of six weeks (T)?
The existing evidence pertaining to the PICOT topics (PHQ-9, DGS, depression, screening, and primary care) was collected by searching MEDLINE, Cochrane, CINAHL, and PubMed databases, which are the most common choice for modern healthcare-related research (Polit & Beck, 2017). Recent literature (not older than seven-eight years) was selected due to the same considerations.
Synthesis of the Evidence
With the above-established challenges in healthcare (Espinoza & Unützer, 2015), as well as the widespread nature of depression (American Psychiatric Association, 2017), depression screening among older adults is clearly a significant clinical problem. As a result, it is critical to identify a tool that can be used for such screening. Specifically for geriatric patients, the GDS has been developed, which has been tested with older adults to demonstrate its validity and reliability. According to a study by Conradsson et al. (2013), its correlation with the Mini-Mental State Examination indicated no significant differences, and the Cronbach’s alpha was ranging between 0.6 and 0.8. As a result, the authors suggested that there was some evidence to the usefulness of the scale, although more research would be required for ensuring that the conclusion was applicable to different groups within the older population.
A significantly better-researched scale is the PHQ-9 tool, as pointed out by El-Den et al. (2017). The author conducted an overview of literature between 1995 and 2015, showing that 14 studies over that period had been dedicated to PHQ-9, proving its validity and reliability (Cronbach’s alpha up to 0.89). In a similar study that focused on a meta-analysis of 40 studies, Mitchell, Yadegarfar, Gill, and Stubbs (2016) demonstrated that PHQ-9 was a workable tool for initial screening with a decent specificity, although its sensitivity was lower than that of a shorter version of the same tool. Finally, a meta-analysis by Moriarty et al. (2015) demonstrated that PHQ-9 was better-suited for primary care settings due to its high specificity and lower sensitivity.
To summarize, PHQ-9 is a good tool for initial screening in primary care that is exceptionally well-studied. However, it is not meant specifically for older adults, which, given the importance and challenges of screening for depression in this population (Espinoza & Unützer, 2015; Zaccagnini & White, 2014), may be a problem. On the other hand, GDS is meant specifically for older adults, but it is not as well-studied. Thus, when older adults are considered, both tools might be a good solution, which calls for their use in the present project.
Application to Practice
Based on the presented evidence, it is reasonable to carry out a training that would assist practitioners in the use of GDS-30 and PHQ-9 for screening depression. This way, the significant problem of the lack of screening in primary care, which may be attributable to the lack of training, will be addressed through evidence-based and (in the case of GDS) tailored solutions. Their application in the present study may also contribute more data on their usability with older geriatric patients.
EBP Action Plan
The goal of this project is to carry out training for the practitioners of the selected site and determine its effect on the number of patients referred to mental health evaluation based on the results of the application and interpretation of PHQ-9 and GHS. Additionally, the applicability of both tools to geriatric patients was considered an objective.
The project involves training, which is why the main population of interest is primary care providers. The fact that primary care providers require training in the field of depression screening is supported by the importance of this procedure (American Psychiatric Association, 2017). Additionally, the study involved patients, specifically older adults (65 years or older). The primary justification for that is that depression screening among older adults is more complicated than among younger people (Espinoza & Unützer, 2015).
Primary healthcare settings were chosen for the project, with a center that the researcher had access to providing permission to carry out the project. Additionally, the managers of the center were helpful in assisting with the education section of the project. The stakeholders, therefore, included the providers, patients, and managers of the stated center. Given that the tools (PHQ-9 and GDS) are meant for preliminary screening that can identify the need for a referral for a mental health evaluation, the selection of the setting is justified.
Kurt Lewin’s Change Model is very often used in nursing practice, and it has been applied in the described effort (Ellis & Abbott, 2018; Spear, 2016). Specifically, the unfreezing stage included identifying existing practices and providing critical information via the training; the change involved the practitioners applying the tools according to their new knowledge. The refreezing could not be carried out within the limited amount of time that was provided for the project, but it has been started due to the participants becoming better equipped to handle the tools correctly.
After their recruitment, the nine practitioners who were involved were requested to apply the screening tool PHQ-9 to fifteen recruited patients (65 years or older). Following that, the participants were subjected to training, which included modules on depression, the importance of screening and its application, and the two tools and their appropriate use. After six weeks, the participants were asked to apply the GHS to the same patients. The data were de-identified and transferred to the researcher according to the below-described data collection procedures.
Based on the research question, the dependent variable is the number of patients referred for mental health evaluation based on their GDS-30 and PHQ-9 evaluations. The questionnaires were the primary tools of data collection as a result; their validity and reliability are demonstrated by the above-cited literature (Conradsson et al., 2013; El-Den et al., 2017). Additionally, some demographic information was collected about the practitioners with the help of a short questionnaire, which included questions about their position and experience. The results of the data analysis are presented below.
All the data were analyzed statistically; descriptive statistics were used to report the key outcomes (Excel was employed for that purpose), and inferential statistics were used to test for correlations between datasets (SPSS was employed for that purpose). The demographics of the participants can be summarized as follows. First, the majority of them were nurses (four people; 45%) followed by nurse practitioners (three people; 33%) and nurses (two people; 22%), as can be seen in Figure 1. Moreover, 56% (five people) reported having some experience with both tools; two people had experience with PHQ-9 only, and two (22%) people reported no experience with either of the tools. Thus, the majority of the participants had had some experience in working with the tools prior to participating in the research.
While the total number of patients that were supposed to be recruited amounted to 15, only nine of them submitted PHQ-9, and six weeks later, only seven completed GDS. Before the training, one patient out of nine was referred (11%). After the training, 42.9% of the seven patients were referred (see Figure 2). In other words, the number of the referrals increased dramatically.
The total scores for PHQ-9 and GHS are presented in Figures 3 and 4. The histograms clearly show that the distribution of the data was abnormal, which prevented the use of parametric tests. As a result, to check for correlations between the data, Kendall’s tau and Spearman’s rho were used (Polit & Beck, 2017). As can be seen from Table 1, no significant correlations between the two datasets were found. Since the results do not correlate very much, the differences in the provided datasets are not unsubstantial.
Table 1. Nonparametric Correlation Test Results.
|Kendall’s tau||PHQ9||Correlation Coefficient||1.000||.441|
|Spearman’s rho||PHQ9||Correlation Coefficient||1.000||.496|
Additionally, it should be pointed out that due to the incredibly small sample (which was the result of the coronavirus pandemic), parametric tests were not applicable to the collected data (Hollander, Wolfe, & Chicken, 2013; Polit & Beck, 2017). Thus, the project used a non-parametric test that could be applied to dependent groups to determine if a statistical difference could be found between the numbers of referrals before and after the training. The results of the sign test, which is an appropriate nonparametric option for data that do not fit other nonparametric alternatives for pared t-test (Hollander et al., 2013), suggest that no statistically significant difference can be found within the presented data (see Table 2). Indeed, with p greater than 0.05, the results of this study cannot be used to claim that the training has affected the number of referrals in a statistically significant way.
|Table 2. Sign Test Results|
|GDS01 – PHQ901|
|Exact Sig. (2-tailed)||.500|
|Exact Sig. (1-tailed)||.250|
Human Subjects Protection
The Institutional Review Board (IRB) was important for the research because of the human subjects involved (Polit & Beck, 2017). It should be pointed out that the risks of participation for participants and patients were minimal; the former only received some additional education, which would not be expected to have negative effects on their careers; the latter were subjected to assessments that would, at most, lead to their referral for reevaluation. As a result, it can be suggested that the risks were almost nonexistent.
Still, all the participants and patients were provided with information about the project, and their informed consent was solicited. Their data were protected through de-identification (the removal of identifying data and its substitution with numbers and roles, for example, Patient 1). Additionally, the hard copies of the information will not be provided to anyone but the provider who handled them. All the data that were provided to the researcher will be kept by a secure, password-locked computer for three years and then destroyed. No hard copies were provided to the researcher.
The organizational factors that were involved in the project were facilitators. The permission to carry out the study was obtained, and the participants were enthusiastic about the training opportunity. However, a major issue was the coronavirus pandemic, which hindered the data collection efforts. Some of the participants still managed to provide their information, which proves their enthusiasm and willingness to assist with the project. Overall, no significant organizational barriers were encountered.
The project was limited by its very small sample, which was further diminished in size by the coronavirus pandemic. Eventually, only seven patients were able to complete GHS. The findings have to be considered from this perspective: they do not imply that the training has had any statistically significant effects, but they also cannot be used to claim the opposite. In the end, more research is required to test the presented training, as well as PHQ-9 and GHS, with a more diverse and larger sample and in other settings (Polit & Beck, 2017), which should help to advance the findings’ generalizability.
Implications for Practice
Based on its data, the study cannot be used to claim that either of the tools is a valid (or more valid) instrument for older adults, and it cannot imply that the presented training has increased the number of referrals in a statistically significant way. Furthermore, the impact of those referrals on patient health and well-being cannot be demonstrated. However, the study did indicate that some of the providers (22% of the sample) might lack the knowledge about commonly used screening tools and can benefit from related training. With the major limitations that are, among other things, the result of the coronavirus, it may be logical to carry out additional similar studies to further test the presented training and its outcomes.
This report details the results of a study dedicated to training practitioners to use PHQ-9 and GDS with older adults in primary care settings. The justification of the project is that the condition is very widely spread but remains underdiagnosed and that in older people, diagnosing depression is difficult. The project involved nine practitioners in a series of training sessions aimed to improve their ability to use PHQ-9 and GDS; 22% of the sample did not know enough about the two tools to apply them. Before the training, the participants referred only one patient, and after the training, three patients were referred.
However, the small sample meant that the data could not demonstrate a statistically significant difference. Additionally, no correlation between the scores of the two tools was found, which is why no evidence of the two tools’ validity was supplied by the project. Still, the research shows that the lack of training may indeed affect primary care professionals, which calls for a more extensive project with a larger sample when the coronavirus restrictions are lifted.
Al-Qadhi, W., Rahman, S., Ferwana, M.S., &Abdulmajeed, I. A. (2014). Adult depression screening in Saudi primary care: Prevalence, instrument and cost. BMC Psychiatry,14(1),1-9. Web.
American Psychiatric Association. (2017). Depression screening rates in primary care remain low. Web.
Cohen, B. E., & Edmondson, D., I. M. (2015). State of the art review: Depression, stress, anxiety, and cardiovascular disease. American Journal of Hypertension,28 (11). 1295-1302. Web.
Conradsson, M., Rosendahl, E., Littbrand, H., Gustafson, Y., Olofsson, B., &Lövheim, H.(2013). Usefulness of the geriatric depression scale 15-item version among very old people with and without cognitive impairment. Aging & Mental Health, 17(5), 638-645. Web.
El-Den, S., Chen, T. F, Gan, Y. L.,Wong, E., & O’Reilly, C. L. (2018). The psychometric properties of depression screening tools in primary healthcare settings: A systematic review. Journal of Affective Disorders,225,503-522. Web.
Ellis, P., & Abbott, J. (2018). Applying Lewin’s change model in the kidney care unit: movement. Journal of Kidney Care, 3(5), 331-333.
Espinoza, R. T., & Unützer, J. (2015). Diagnosis and management of late-life unipolar depression. Web.
Farrer, L. M., Gulliver, A. G., Bennett, K., Fassnacht, D. B., & Griffiths, K. M. (2016). Demographic and psychosocial predictors of major depression and generalised anxiety disorder in Australian university students. BMC Psychiatry, 16, 241. Web.
Fekadu, N., Shibeshi, W., &Engindawork, E. (2017). Major depressive disorder: Pathophysiology and clinical management. Journal of Depression and Anxiety, 6(1), 255-261. Web.
Ferenchick, E. K. (2019). Depression in primary care: Part 1—screening and diagnosis. BMJ, 365, 1-10. Web.
Hollander, M., Wolfe, D. A., & Chicken, E. (2013). Nonparametric statistical methods. New York, NY: John Wiley & Sons.
Mitchell, A. J., Yadegarfar, M., Gill, J., & Stubbs, B. (2016). Case finding and screening clinical utility of the patient health questionnaire (PHQ-9 and PHQ-2) for depression in primary care: A diagnostic meta-analysis of 40 studies. BJPsychOpen, 2(2), 127-138. Web.
Moriarty, A., Gilbody, S., McMillan, D., & Manea, L. (2015). Screening and case finding for major depressive disorder using the patient health questionnaire (PHQ-9): A meta-analysis. General Hospital Psychiatry, 37(6). 567-576. Web.
National Institute of Mental Health. (2019). Major depression. Web.
Park, M., &Unützer, J. (2011). Geriatric depression in primary care. Psychiatric Clinics of North America, 34(2), 469-487. Web.
Pfizer Inc. (1999). Patient Health Questionnaire. Web.
Polit, D.F., & Beck, C.T. (2017). Nursing research: Generating and assessing evidence for nursing practice (10th ed.). Philadelphia, PA: Lippincott, Williams & Wilkins.
Spear, M. (2016). How to facilitate change. Plastic Surgical Nursing, 36(2), 58-61.
Yesavage, J., Brink, T., Rose, T., Lum, O., Huang, V., Adey, M., & Leirer, V. (1982). Development and validation of a geriatric depression screening scale: A preliminary report. Journal of Psychiatric Research, 17(1), 37-49. Web.
Zaccagnini, M. E., & White, K. W. (2014). The doctor of nursing practice essentials: A new model for advanced nursing practice. New York, NY: Jones and Bartlett Publishers.