Construction and Validation of a Risk Warning Model of Depression in Patients with Pulmonary Nodule

Background . To establish a depression risk warning model for patients with pulmonary nodules and to provide a theoretical basis for medical staf to identify high-risk patients early and quickly and take timely intervention measures. Methods . A total of 535 hospitalized patients with pulmonary nodules were selected, and the relevant data were analyzed by single-factor analysis. Binary logistic regression analysis was used to determine the independent risk factors for depression in patients with pulmonary nodules and to establish a risk warning model. Te Hosmer − Lemeshow test and receiver operating characteristic (ROC) curve were used to evaluate the goodness of ft and prediction efect of the model, and the cross-validation method was used to verify the efcacy of the model. Results . Te prevalence of depression in patients with pulmonary nodules was 47.29%. Univariate analysis showed that CRP, albumin, creatinine, phosphorus, calcium, triglyceride, cholesterol, low-density lipoprotein, high-density lipoprotein, β 2-microglobulin, objective support, social support, and education level were related to depression in patients with pulmonary nodules ( P < 0 . 05 ) . Binary logistic regression analysis showed that serum CRP, calcium, social support, and education level were independent risk factors for depression in patients with pulmonary nodules. Te area under the ROC curve/sensitivity/specifcity of serum CRP, calcium, social support, and education level was 0.78/86.88%/80.65%, 0.79/84.40%/75.59%, 0.83/89.91%/80.22%, and 0.81/85.96%/79.19%, and the accuracy of cross-checked risk warning model was 84.97%. In the Hosmer − Lemeshow test, P � 0.926, area under the ROC curve was 0.98, sensitivity was 98.17%, and specifcity was 93.55%. Te accuracy of the cross-checked risk warning model was 84.97%, indicating that the prediction efect of the model was good. Conclusions . Serum CRP, calcium, social support, and education level are the independent risk factors of depression in patients with pulmonary nodules, and the risk warning model based on them has a good early warning efect on depression in patients with pulmonary nodules. Te risk warning model established in this study has a good predictive efect on depression in patients with pulmonary nodules.


Introduction
Nowadays, with the advancement of people's living standards, the continuous improvement of health awareness, and the widespread use of low-dose computed tomography (LDCT), pulmonary nodules are most discovered during physical examination [1].By chest radiography or chest CT examination, most pulmonary nodules are difcult to distinguish between benign and malignant, and such patients usually require outpatient follow-up for more than 2 years to observe changes in the nodules and determine whether further treatment is required [2,3].Most pulmonary nodules do not progress to lung cancer, but because of the frst discovery and long-term follow-up, patients may encounter signifcant stress, adverse emotional changes, and distress.A high level of illness uncertainty reduces patients' quality of life and treatment efect and has a signifcant impact on their prognosis [4,5].According to studies [6,7], 40% of patients with pulmonary nodules experience anxiety and depression, whereas this proportion is only 7% in healthy physical examiners who do not have pulmonary nodules.Anxiety and depression symptoms in patients with pulmonary nodules can lead to immune dysfunction and infammation [8,9].As a result, it is critical to actively seek predictive tools that correspond to the depressive characteristics of patients with pulmonary nodules.Trough a case-control study, this study investigated the risk factors associated with pulmonary nodules depression and developed a risk early warning model to help the medical staf determine the risk of depression in patients early and quickly, early identifcation of high-risk groups and nursing intervention.

Objects Participants and Methods
2.1.Research Objects Participants.Patients with pulmonary nodules who were admitted to Cardiothoracic Surgery at Jiangnan University's Afliated Hospital between January 1, 2020, and October 10, 2021, were chosen as research subjects.Te following were the inclusion criteria: ① patients with pathological diagnosis of pulmonary nodule; ② Age ≥18 years; ③ Understand the disease; ④ Have normal language expression ability and comprehension ability and agree to participate in the study.Exclusion criteria: ① have a personality disorder or patients already using psychotropic medications; ② complicated with serious cardiovascular, liver, and kidney diseases or malignant tumors; ③ patients diagnosed with mental illness before illness; ④ Patients who had participated in psychological intervention in the past 3 months.Informed consent was obtained from all patients.

Recruitment and Data
Collection.After selecting our research subjects according to our inclusion and exclusion criteria, we had one-on-one communication with the research subjects to explain our research content and the signifcance of our research.According to the Helsinki Declaration, verbal informed consent of the participants was obtained before the start of our study.Patients were evaluated using the self-rating depression scale, and patients with a score lower than 50 were defned as the nondepression group, and those with a score higher than 50 were defned as our depression group.
Te patients' general information background information was gathered, which included: gender, age, education level, monthly household income per capita, profession, patient source, type of health care, drinking, smoking, marriage, total score for social support, these data were obtained through interviews with patients; family history of mental illness, the history of surgery, the history of chronic illness, tumor related history, staging, number of chemotherapies, whether the transfer, serum CRP, albumin, serum creatinine, phosphorus, calcium, β2 microglobulin, triglycerides, total cholesterol, tow-density lipoprotein, high-density lipoprotein, these data were obtained by reviewing hospital medical records and obtaining consent from the hospital feld.Before data collection: two researchers who had received standardized training explained the meaning of scale items and notes for flling out the scale to patients one by one using unifed instructions.Adopt efective communication methods to ensure the accuracy of data.Data collection and entry: At least two researchers must be present at the same time to ensure the data's validity and completeness.After data collection: Data cannot be changed for any reason.Efective means of communication have been employed to ensure the accuracy of the data.Data collection and entry: At least two investigators have been present at the same time to ensure data validity and completeness.After data collection: Data have not been changed for any reason.Convenience sampling was used in this study.In this study, 550 questionnaires were distributed, 545 were recovered, 10 invalid questionnaires were eliminated, and 535 valid questionnaires were obtained, yielding an efective rate of 97.73%.

Instruments
(1) Te Self-Rating Depression Scale (SDS) was developed by Zung in 1965 [10,11], and SDS includes 20 items that assess symptoms of depression.Participants rated each item based on how they felt during the previous week.Item responses were scored on a four-point rating scale (1-4), a higher SDS score indicates a higher level of depression.Te sum of the scores of the 20 items was the total score, which was multiplied by 1.25 to obtain the SDS index score.Te Chinese version of SDS was published in the late 1980s.According to the Chinese SDS Manual [12], people with a score of ① 50-59 were classifed as mild depression; ② 60-69 were classifed as moderately to signifcantly depressed; ③ Individuals with scores of 70 and above were classifed as having severe to extreme depression.In the present study, Cronbach's alpha was 0.81.(2) Te Social Support Rating Scale (SSRS) is used to assess the perception of social support and satisfaction with this social support, which was developed by Xiao [13,14] according to the environmental and cultural conditions in China.It consists of 10 items, including objective support (3 items), subjective support level (4 items), and social support utilization (3 items).Items 1-4 and 8-10 were scored on a fourpoint scale (1-4 points).Item 5 was divided into fve sub-questions, and each sub-question was graded from "none" to "full support" on a 1-4-point scale.
For items 6 and 7, points were given according to the number of sources in the "following sources."If the answer was "no source," the score was 0. Te higher the total score, the higher the overall social support.Te score was 0. Te higher the total score, the higher the overall social support.≥45 was defned as high support [15].Te scale has been widely used in the Chinese population and has been shown to have high reliability and validity [16,17].In the present study, the sums of total scale and three subscales were calculated.Te Cronbach's α value of the scale, 0.86, indicated good internal consistency.

Statistical Analysis.
For data analysis, SPSS 25.0 was used.We used the χ 2 tests and the logistic regression analysis.To assess the predictive value of each independent index and the risk, early warning model, on the depression of pulmonary nodule patients, the receiver operating characteristic (ROC) curve was used.Te risk early warning model's efciency was tested using the leave-one-out crossvalidation method.1.

Te Status of Depression in
Te risk warning model formula was developed by using the above variables B as coefcients in the logistic regression model.P � 1/{1 + exp[−(−0.03× serum CRP + 1.32 × calcium + 0.42 × social support−4.20× education level (middle school)−3.00× education level (high school and above)−30.59)]}.Te predicted probability value was calculated using the predicted probability as the diagnostic variable and the occurrence of depression as the status variable.Te results showed that the area under the ROC curve of the risk early warning model was 0.98, the cut-of value was 0.43, and the sensitivity and specifcity were 98.17% and 93.55%, respectively.As shown in Figure 2. Te result of the Hosmer-Lemeshow deviation test was χ 2 � 0.462, P � 0.926, indicating that the model was wellcalibrated.

Efciency Test of Risk Warning Model.
Taking depression as a categorical variable and the predicted probability value as an independent variable, the consistency between the predicted depression grouping and the measured depression grouping of the warning model was tested.Te results showed that the classifcation accuracy of the risk warning model was 84.97%.As shown in Table 3.

Statement of Principal Findings. Depression can hurt
patients' quality of life as well as the disease's therapeutic efect.Patients with pulmonary nodules often struggle to distinguish between benign and malignant pulmonary nodules, and because of their lack of understanding of the disease and their fear, such patients frequently sufer from major depression [18].Tis study discovered that patients with pulmonary nodules had a high level of depression, with a 47.29% incidence.Tis study's depression risk prediction model confrmed that serum CRP, calcium, sleep quality, social support, education level, and quality of life all had a direct predictive efect on depression.

Risk Factors of Depression in Patients with Pulmonary
Nodules.Infammation caused by pulmonary nodules and depression is linked.Infammation causes more severe depression and a lower likelihood of surviving [19].CRP is an acute phase response protein that can be indicative of congenital infammation.CRP has been given the most attention and has been linked to depression [20].CRP and proinfammatory cytokines [21,22] are two biomarkers that have been linked to depression.
Because the diet is a modifable risk factor for depression, dietary changes may reduce the burden of depression [23].Adopting healthy dietary patterns that correspond to foodbased dietary recommendations and nutritional needs is critical for preventing, slowing, or controlling the progression of depressive symptoms and promoting optimal mental health [24].
Emotional disorders are frequently associated with sleep disorders.Sleep disorders are not only the most common pre-depression symptoms, but they can also predict the occurrence and outcome of depression [25].Suicidal ideation and attempt are independent predictors of sleep disorder [26].At the same time, depression can exacerbate sleep problems, and sleep and depression have a two-way predictive efect [27].
Tere is also a link between the degree of depression and the quality of life [28].Pulmonary nodule patients have a lower quality of life, and most of them experience negative emotions, primarily anxiety and depression.Te more severe the depression, the lower the quality of life of pulmonary nodule patients [29].By improving patients' psychological adaptation level, social support can alleviate Perspectives in Psychiatric Care depressive symptoms in pulmonary nodule patients [30], obtain more social support, and efectively relieve anxiety and depression in pulmonary nodule patients [31].Perceived social support assists pulmonary nodules patients in developing efective coping strategies to cope with the disease.Perceived social support can help patients not only reduce their economic burden but also establish confdence in overcoming the disease and cultivate a positive and optimistic attitude, which is conducive to eliminating negative emotions and promoting disease improvement [32].
Tere are diferences in the prevalence of depression among pulmonary nodules patients with varying levels of education.Patients with a high education level can learn in a variety of ways, have a clear and scientifc understanding of their disease, communicate more smoothly with medical staf, and reduce their psychological burden; patients with a low education level have fewer available social resources and are more likely to sufer from depression [33].Such patients should be given special attention in clinical settings.

Strengths and Conclusions
Te risk prediction model of depression in patients with pulmonary nodules developed in this study has good prediction efciency, and the data are convenient and easy to obtain, making it useful for risk assessment and depression prevention in patients with the pulmonary nodule.Clinical staf should conduct early assessment and intervention based on the factors that infuence depression, improve information mastery in patients with pulmonary nodules, and reduce the occurrence of depression in patients with pulmonary nodules.

Figure 1 :
Figure 1: ROC curve of serum CRP, calcium, social support, and education level.

Figure 2 :
Figure 2: ROC curve of the risk warning model of depression in patients with pulmonary nodules.

Table 1 :
Univariate analysis of depression in patients with pulmonary nodule.

Table 2 :
Binary logistic regression analysis of depression in patients with pulmonary nodules.

Table 3 :
Efciency test of risk warning model.

6
Perspectives in Psychiatric Care 4.3.Implications for Policy, Practice and Research.Due to the high incidence of depression in patients with pulmonary nodules and its negative consequences, there is an urgent need for a rapid and operable method to conduct risk screening for such patients, but there are few similar studies now.Te risk warning model of depression in patients with pulmonary nodules developed in this study consisted of six components: serum CRP, calcium, sleep quality, social support, education level, and quality of life.Te area under the ROC curve was 0.98, which was greater than the predictive efcacy of independent risk factors, and the diagnostic sensitivity and specifcity were respectively 98.17% and 93.55%.Te diagnostic value outperforms a single index.After cross validation, the prediction accuracy of the model was 85.92%.In general, the model has a good predictive ability for the early judgment of depression risk of patients, and medical staf can put forward more appropriate nursing measures according to the model in the follow-up work to improve the treatment efect of patients.To summarize, serum CRP, calcium, sleep quality, social support, education level and quality of life are independent risk factors for depression in patients with pulmonary nodules.Te risk warning model developed because of this is highly operable and reliable, with superior diagnostic efcacy when compared to a single indicator.It can be used to quickly determine whether patients with early pulmonary nodules are at risk for depression.