- Email: [email protected]
History of Depression in Lung Cancer Patients: Impact of Delay
Clinical Oncology xxx (2017) 1e8 Contents lists available at ScienceDirect
Clinical Oncology journal homepage: www.clinicaloncologyonline.net
Original Article
History of Depression in Lung Cancer Patients: Impact of Delay M. Iachina *y, M.M. Brønserud zx, E. Jakobsen zx, O. Trosko {, A. Green x * Center
for Clinical Epidemiology, Odense University Hospital, Odense, Denmark The Institute of Clinical Research, University of Southern Denmark, Odense, Denmark z Department of Thoracic Surgery, Odense University Hospital, Odense, Denmark x OPEN, Odense Patient Data Exploratory Network, Odense University Hospital/Department of Clinical Research, University of Southern Denmark, Odense, Denmark { Research Unit of Psychiatry, Department of Clinical Research, University of Southern Denmark, Odense, Denmark y
Received 6 September 2016; received in revised form 20 March 2017; accepted 21 March 2017
Abstract Aims: To examine the influence of a history of depression in the process of diagnostic evaluation and the choice of treatment in lung cancer. Materials and methods: The analysis was based on all patients with non-small cell lung cancer who were registered in 2008e2014; in total, 27 234 patients. To estimate the effect of depression on the diagnostic process and the choice of treatment in lung cancer we fitted a logistic regression model and a Cox regression model adjusting for age, gender, resection and stage. Results: Depression in a patient’s anamnesis had no significant effect on the delay in diagnostic evaluation (hazard ratio ¼ 0.99 with 95% confidence interval 0.90; 1.09). Patients with a history of periodic depression had a 33% lower treatment rate (odds ratio ¼ 0.66 with 95% confidence interval 0.51; 0.85) than patients without a history of depression. Conclusions: Our study shows that patients with a history of periodic depression need special attention when diagnosed with lung cancer. Ó 2017 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved.
Key words: Depression; diagnostic evaluation; lung cancer; socio-economic status; treatment
Introduction With an annual incidence of 4500 patients, primary lung cancer is one of the most common cancers in Denmark. In general, the lung cancer prognosis is poor, with a 5 year survival of about 10e12%. Comorbidity is common in patients with lung cancer (26e81%) [1] and an increased comorbidity increases mortality significantly [2]. Depression has been associated with increased mortality in patients with lung cancer [3,4], but not much is known as to the reason. Spiegel and Giese-Davis [5] identified several reasons why depression may have an effect on the mortality of cancer patients in general. First, depression may have a pathophysiological effect via a neuroendocrine and
Author for correspondence: M. Iachina, Center for Clinical Epidemiology, Odense University Hospital, Sdr. Boulevard 29, Entrance 216, Ground Floor East, DK-5000 Odense C, Denmark. Tel: þ45-21158110. E-mail address: [email protected] (M. Iachina).
immunological function that influences mortality. Second, many of the symptoms of cancer and the side-effects of its treatment are similar to those of depression, e.g. sleep disturbance, anorexia, fatigue and concentration difficulties. Therefore, referral to an examination for cancer can be delayed for depressed patients. Third, patients with depression may be less likely to adhere to preventive screening procedures, cancer treatments or recommendations for maintaining good health. Depression and smoking are associated [6] and lung cancer-specific symptoms, such as cough and shortness of breath, are common in heavy smokers, thus hiding the lung cancer symptoms and adding further delay to the diagnosis. Depressed patients may miss screening appointments and are thus more likely to delay the treatment and less likely to receive treatment or to implement treatment [7]. Diagnostic procedures in suspected lung cancer are carried out to establish a diagnosis and the stage of lung cancer. The choice of treatment is furthermore based on the patient
http://dx.doi.org/10.1016/j.clon.2017.03.014 0936-6555/Ó 2017 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved.
Please cite this article in press as: Iachina M, et al., History of Depression in Lung Cancer Patients: Impact of Delay, Clinical Oncology (2017), http://dx.doi.org/10.1016/j.clon.2017.03.014
2
M. Iachina et al. / Clinical Oncology xxx (2017) 1e8
performance and comorbidity. Patients may be treated with surgical resection or chemo- and/or radiotherapy. Recently targeted and immunotherapy are added to treatment options. Surgical resection of the tumour is associated with the best survival, but only about 20% of the patients are eligible for resection. The clinical stage and performance are the most important factors when deciding on the treatment. A range of other prognostic factors are also taken into account before the final decision about treatment is made, including age, smoking, alcohol or drug abuse, patient preferences and comorbidity. The aim of this study was to examine the influence of the history of depression in the process of diagnostic evaluation and the choice of treatment in lung cancer. A similar study has previously been carried out in the USA by Sullivan et al. [4], but without taking into account the differences in the patient’s socio-economic status. They found no association of depression diagnosis with receiving treatment. It is well known that there is a strong association between socioeconomic position and the experience of depression [8] and that a low socio-economic status is associated with late-stage cancer diagnoses and the type of treatment received [9,10]. Here we investigated the effect of a history of depression when adjusting for the patient’s socioeconomic status.
Materials and Methods We used data from the four national registries: The Danish Lung Cancer Registry We used the Danish Lung Cancer Registry (DLCR) to identify patients in Denmark with lung cancer. Since 2000 the DLCR has monitored and evaluated the quality of treatment effort concerning lung cancer [11]. All Danish lung cancer patients are included in the DLCR and to this date the database contains data on more than 55 000 cases of lung cancer. Between 2000 and 2002 clinicians identified and reported new patients to the DLCR, but since 2003 lung cancer patients have been identified in the Danish National Patient Registry (DNPR), where the first occurrence of the diagnostic codes DC34 and DC33, according to the International Classification of Diseases 10 (ICD-10), are identified. Information about these patients and their activities (procedures and treatments) registered with the DNPR, together with information from the Danish Pathology Registry, provides the basis of establishing patient trajectories and associated diagnostic procedures and treatments in the DLCR. Only about 75% of the information required for the DLCR is available in the Danish central registries and the participating clinicians are requested to supplement with the remaining about 25% of information from their medical records. As participation in this supplementation is mandatory by law, data completeness is very high (more than 95%). Data on comorbidity were obtained from the DNPR [12]. The history and status on depression is not
requested routinely by the DLCR and was obtained specifically for this project. The Danish Depression Database We retrieved data from the Danish Depression Database [13] to identify patients with a history of depression. The database includes all patients who are 18 years or older with a diagnosis of depression, have permanent residence in Denmark and are hospitalised or affiliated with a psychiatric hospital in Denmark. The disease is diagnosed according to ICD-10 criteria and includes all sub-codes: DF32.X - depressive episode (DF32.0; DF32.1; DF32.2; DF32.3; DF32.8; DF32.9); DF33.X e recurrent depressive disorder (DF33.0; DF33.1; DF33.2; DF33.3; DF33.4; DF32.8; DF32.9); DF34.1X e dysthymia; DF06.32 e organic affective disorder, depressive. We included both in-patients and out-patients who had at least one hospital contact with depression (DF32.X depressive episode and DF33.X recurrent depressive disorder) within 10 years before the start of the lung cancer diagnostics process. We excluded from our analysis patients with dysthymia and organic depression as the definition and diagnostic delineation of these two groups is more uncertain. Here we will examine the effect of the following three groups of patients: one includes patients with the diagnosis ‘depressive episode’ (ICD-10 DF32.X), another includes patients with the diagnosis ‘recurrent depressive disorder’ (ICD-10 DF33.X) and the last one includes patients from both groups. The National Patient Register We included information on comorbidity for each patient up to 10 years before the lung cancer diagnosis, using the DNPR, which was established in 1977. This register contains data on all interventions related to diagnostic evaluation and treatment for somatic patient admissions in Denmark [14]. We used a modification of the Charlson comorbidity index (CCI) for the classification of somatic comorbidity [15]. All hospital contacts represented with a cancer diagnosis, registered within 150 days before the date of lung cancer diagnosis, were excluded from contribution to the cancer comorbidity group in order to avoid the influence of misclassification of cases with lung cancer before the final diagnosis. The Income Statistics Register We included information on the socio-economic status of each patient at the time of the lung cancer diagnosis, using information available with The Income Statistics Register [16]. The socio-economic class was defined as household income in the year the patient was diagnosed with lung cancer, adjusted for the number of people in the household. The socio-economic class was divided by the median into two groups, low and high.
Please cite this article in press as: Iachina M, et al., History of Depression in Lung Cancer Patients: Impact of Delay, Clinical Oncology (2017), http://dx.doi.org/10.1016/j.clon.2017.03.014
M. Iachina et al. / Clinical Oncology xxx (2017) 1e8
All permanent residents in Denmark are assigned to a unique personal identification number, which is used in all Danish registers enabling data linkage [15]. Study Population The analysis is based on all patients with non-small cell lung cancer who were registered in 2009e2011, in total 27 234 patients. For each patient, comorbidity information was collected using the DNPR up to 10 years before the lung cancer diagnosis. This register contains data including coding of all interventions related to diagnostic evaluation and treatment for all somatic patient admissions in Denmark. Statistical Analyses For the sake of simplicity, we will in the following refer to the group of patients, who have had depression in the anamneses, as a group of patients with depression and correspondingly the group of patients who do not have depression in the anamneses will be called patients without depression. We used Fisher’s exact test to compare baseline characteristics of patients with and without depression, and to compare patients with a depressive episode to patients without depression, and patients with a recurrent depressive disorder to patients without depression. We used a logistic regression model, with clinical stage as outcome, to see if there was any delay in the referral to examination for lung cancer for depressed patients compared with non-depressed patients. A Cox regression model was used to estimate the effect of depression on the length of the primary investigations, and to estimate the effect of depression on the time from the primary investigation to the start to treatment, for those patients who received specific treatment. Using a logistic regression model we also estimated the effect of depression on the risk of receiving a treatment in general, receiving surgical treatment and receiving oncological treatment. The following variables acted as confounders in the prediction model: age: young age was defined as age lower than the mean age for the patient population; gender; comorbidity: 0, CCI score of 0; 1, CCI score of 1; 2, than CCI score >1; clinical tumour stage when possible. As there is an interaction between socio-economic status and the experience of depression, the socio-economic class was used as a stratification variable. The socio-economic class was taken from the patient’s personal income in the year of lung cancer diagnosis. The personal income for our study population ranged from 95 000 DKK to 668 000 DKK per year with a 25% fractile of 162 000 DKK, a 50% fractile of 200 000 DKK and a 75% fractile of 270 000 DKK. The variable socio-economic class was divided into: low e socioeconomic class being lower than the mean socioeconomic class for the patient population; high e socioeconomic class being equal or higher than the mean socio-economic class for the patient population. For all outcomes we carried out three separate analyses: first to compare patients with depression with
3
patients without depression; second to compare patients with a depressive episode (ICD-10 DF32) with patients without depression; third to compare patients with a recurrent depressive disorder (ICD-10 DF33) with patients without depression. Moreover, as an association between socio-economic status and depression is not unequivocal we also completed analyses stratified by socio-economic class. Ethical Aspects DLCR is approved as a national clinical database by the Danish National Board of Health (j.nr. 7-201-03-15/1). Permission for the present study was granted by the Danish Data Protection Agency (j.nr. 2008-58-0035).
Results In total we extracted 27 234 patients from the DLCR who had no hospital contact with diagnosed depression within 10 years before the primary lung cancer diagnosis and 508 who had hospital contact (see Table 1); 215 of those had a depressive episode, 293 had a recurrent depressive disorder. Patients with depression were younger and with a preponderance of females compared with patients without depression. Similar patterns of age and gender distribution were observed in patients with a depressive episode compared with patients with a recurrent depressive disorder. More patients from the depression group had a high comorbidity (42.5%) compared with the group of patients without depression (31.5%). Among patients with a depressive episode, 39.1% had a high comorbidity, whereas in patients with a recurrent depressive disorder 45.1% had a high comorbidity. There was no difference in the stage and socio-economic distribution for patients with depression regardless of depression group. There was no difference in the duration of the diagnostic process between all patient groups. Table 2 shows estimates of the univariate analyses. There was no difference in the stage distribution between patients with depression (53.9% patients with high stage) and without depression (57.3% patients with high stage, P ¼ 0.312). The same applies to the patients from the depressive episode group (53.5% patients with high stage) as well as to the patients from the recurrent depressive disorder group (54.3% patients with high stage). The results suggest that in the high socio-economic group patients with depression had a lower chance (51.0%) of having a high stage compared with the patients without depression (58.2%, P ¼ 0.037). There was no difference in the duration of the diagnostic evaluation process or in the days from the primary investigation to treatment among depression groups. The ratio for treatment was higher for patients without depression (75.4%) than for patients with depression (72.1%); 77.2% of the patients with a depressive episode received treatment, whereas the treatment ratio was only
Please cite this article in press as: Iachina M, et al., History of Depression in Lung Cancer Patients: Impact of Delay, Clinical Oncology (2017), http://dx.doi.org/10.1016/j.clon.2017.03.014
0.106 0.753
72 (24.6%) 133 (45.4%) 88 (30.0%)
0.312
0.227
8714 (32.6%) 15 310 (57.3%) 2702 (10.1%)
8714 (30.2%) 11 313 (42.3%) 7332 (27.4%)
64 (29.8%) 96 (44.7%) 55 (25.6%)
0.576 0.527
103 (35.1%) 159 (54.3%) 31 (10.6%)
<0.000
77 (35.8%) 115 (53.5%) 23 (10.7%)
<0.000 0.053
105 (35.8%) 56 (19.1%) 132 (45.1%) 12 304 (46.0%) 5998 (22.4%) 8424 (31.5%)
91 (42.3%) 40 (18.6%) 84 (39.1%)
<0.000 <0.000 293 168 (57.3%) 97 (33.1%) <0.000 0.027 215 125 (58.1%) 95 (44.2%) <0.000 <0.000 26 726 12 231 (45.8%) 13 838 (51.8%)
Number 508 Age (young) 293 (57.7%) Gender (male) 192 (37.8%) Comorbidity 0 196 (38.6%) 1 96 (18.9%) >1 216 (42.5%) Stage 0, I, II, IIIa 180 (35.4%) IIIb, IV 274 (53.9%) Missing 54 (10.6%) Socio-economic status Low 136 (26.8%) High 229 (45.1%) Missing 143 (28.2%)
With depression
Without depression
P value for test depression versus none
Depression episode
P value for test depression episode versus none
Recurrent depressive disorder
P value for test recurrent depressive disorder versus none
M. Iachina et al. / Clinical Oncology xxx (2017) 1e8 Table 1 Baseline characteristics of patients with depression and without depression, as well as of patients with a single depression episode and patients with a history of periodic depression
4
68.3% for the patients with a recurrent depressive disorder. The difference in the treatment ratio between patients with a recurrent depressive disorder and patients without depression was statistically significant (P ¼ 0.005). Only 79.1% of the patients from the high socio-economic group with recurrent depressive disorder received treatment against 84.9% of the patients from the high socio-economic group without depression. There was almost no difference in the ratio for surgical treatment in the patients from the low socio-economic groups (20.1% for the patients without depression, 19.9% for the patients with depression, 20.3% for the patients with a depressive episode and 19.4% for the patients with a recurrent depressive disorder). Patients from the high socio-economic group had a substantially higher ratio for surgical treatment (22.5% for the patients without depression, 27.5% for the patients with depression, 26.0% for the patients with a depressive episode and 28.6% for the patients with a recurrent depressive disorder). The patients with depression received statistically significantly less oncological treatment than the patients without depression (58.5% for the patients without depression versus 51.9% for the patients with depression). There was no difference in the ratio for oncological treatment between the group of patients with a depressive episode (57.7%) and the group of patients without depression (58.5%), but only 47.8% of the patients with the recurrent depressive disorder received oncological treatment. The difference in the oncological treatment ratio between patients with a recurrent depressive disorder and patients without depression was statistically significant (P < 0.001). The difference in the ratio for oncological treatment for the patients from the low socio-economic groups was not significant (59.5% for the patients without depression, 58.1% for the patients with depression, 59.4% for the patients with a depressive episode and 56.9% for the patients with a recurrent depressive disorder). The ratio for oncological treatment was significantly lower (P ¼ 0.015) for patients with depression from the high socio-economic group (54.6%) compared with patients without depression (62.3%). The patients with a recurrent depressive disorder from the high socio-economic group had the lowest ratio for oncological treatment (51.1%) among the patients from the high socio-economic group. Table 3 shows estimates of the multivariate analyses. In each model we used age, gender, comorbidity as a confounder and stage where possible. The adjusted results for the stage confirmed the unadjusted findings shown in Table 2. There was no difference in the stage distribution between patients with depression and without depression (odds ratio ¼ 0.9, confidence interval 0.74; 1.09). The same applies to the patients from the depressive episode group (odds ratio ¼ 0.86, confidence interval 0.65; 1.16) as well as for the patients from the recurrent depressive disorder group (odds ratio ¼ 0.93, 0.72; 1.19). The results indicate that the ratio for a high stage was lower for patients with depression in the high socio-economic group (odds ratio ¼ 0.78, confidence interval 0.59; 1.03) and patients from the high socio-
Please cite this article in press as: Iachina M, et al., History of Depression in Lung Cancer Patients: Impact of Delay, Clinical Oncology (2017), http://dx.doi.org/10.1016/j.clon.2017.03.014
Outcome
Socio-economic class
With depression
Without depression
P value for test depression versus none
Depression episode
P value for test depression episode versus none
Recurrent depressive disorder
P value for test recurrent depressive disorder versus none
Stage (high versus low)
All Low High All Low High
274 (53.9%) 67 (54.5%) 106 (51.0%) 30.1 (3.4) 33.6 (6.9) 35.6 (6.2)
15 310 (57.3%) 4066 (55.5%) 6047 (58.2%) 28.8 (0.5) 31.9 (0.9) 32.8 (0.9)
0.312 0.903 0.037 0.731 0.829 0.668
115 (53.5%) 30 (53.6%) 44 (50%) 25.8 (1.9) 25.9 (2.0) 28.6 (3.6)
0.527 0.779 0.122 0.594 0.592 0.672
159 (54.3) 37 (56.1%) 62 (51.7%) 33.3 (5.7) 40.3 (12.8) 40.7 (10.6)
0.576 0.922 0.155 0.369 0.421 0.3646
All Low High
42.3 (2.1) 42.3 (2.1) 43.4 (40.9)
40.7 (0.4) 44.9 (0.8) 40.9 (0.52)
0.551 0.679 0.509
44.6 (4.2) 45.3 (2.9) 46.3 (7.8)
0.344 0.969 0.343
40.5 (1.6) 39.6 (3.0) 41.1 (1.9)
0.996 0.543 0.997
All Low High All Low High All Low High
366 (72.1%) 106 (77.9%) 188 (82.1%) 102 (20.1%) 27 (19.9%) 63 (27.5%) 264 (51.9%) 79 (58.1%) 125 (54.6%)
20 144 (75.4%) 6436 (79.6%) 9614 (84.9%) 4500 (16.8%) 1625 (20.1%) 2546 (22.5%) 15 644 (58.5%) 4811 (59.5%) 7068 (62.25%)
0.085 0.239 0.277 0.053 0.941 0.073 0.003 0.733 0.015
166 (77.2%) 51 (79.7%) 82 (85.4%) 42 (19.5%) 13 (20.3%) 25 (26,0%) 124 (57.7%) 38 (59.4%) 57 (59.4%)
0.388 0.998 0.901 0.290 0.969 0.406 0.795 0.976 0.533
200 (68.3%) 55 (76.4%) 106 (79.1%) 60 (20.5%) 14 (19.4%) 38 (28.6%) 149 (47.8%) 41 (56.9%) 68 (51.1%)
0.005 0.490 0.091 0.101 0.886 0.100 0.000 0.653 0.008
Days of primary investigation Mean (standard deviation) Days from primary investigation to treatment Mean (standard deviation) Treatment (yes versus no) Resection (yes versus no) Oncological treatment (yes versus no)
M. Iachina et al. / Clinical Oncology xxx (2017) 1e8 5
Please cite this article in press as: Iachina M, et al., History of Depression in Lung Cancer Patients: Impact of Delay, Clinical Oncology (2017), http://dx.doi.org/10.1016/j.clon.2017.03.014
Table 2 Effect of depression (yes versus no) for different socio-economic patient groups on staging process and treatment
6
M. Iachina et al. / Clinical Oncology xxx (2017) 1e8
Table 3 Effect of depression (yes versus no) estimated in the models that describe the primary investigation and the treatment offers; logistic regressions are shown as odds ratio (95% confidence interval), Cox regressions are shown as hazard ratio (95% confidence interval) Outcome
Socio-economic class
Depression versus none
Depression episode versus none
Recurrent depressive disorder versus none
Stage (odds ratio)* (high versus low)
All Low High All Low High
0.90 1.01 0.78 0.99 0.98 0.94
0.86 0.96 0.72 0.92 0.89 0.91
0.93 1.07 0.83 1.05 1.06 0.97
All Low High
0.91 (0.81; 1.01) 0.96 (0.78; 1,17) 0.87 (0.75; 1.01)
0.88 (0.75; 1.03) 0.85 (0.63; 1.13) 0.86 (0.69; 1.08)
0.93 (0.80; 1.07) 1.07 (0.82; 1.41) 0.88 (0.72; 1.07)
All Low High All Low High All Low High
0.79 0.86 0.79 1.16 0.93 1.25 0.76 0.95 0.73
1.02 0.97 0.97 1.13 0.97 1.18 0.94 1.01 0.87
0.66 0.78 0.69 1.18 0.90 1.30 0.65 0.91 0.66
Days of primary investigation (hazard ratio)y Mean (standard deviation) Days from primary investigation to treatment (hazard ratio)y Mean (standard deviation) Treatment (odds ratio)y (yes versus no) Resection (odds ratio)y (yes versus no) Oncologic treatment (odds ratio)y (yes versus no) * y
(0.74; (0.71; (0.59; (0.90; (0.82; (0.82;
(0.64; (0.57; (0.55; (0.93; (0.61; (0.93; (0.64; (0.68; (0.57;
1.09) 1.46) 1.03) 1.09) 1.17) 1.08)
0.96) 1.31) 1.11) 1.45) 1.43) 1.68) 0.91) 1.35) 0.96)
(0.65; (0.56; (0.47; (0.80; (0.68; (0.74;
(0.74; (0.52; (0.55; (0.81; (0.53; (0.75; (0.72; (0.61; (0.57;
1.16) 1.62) 1.09) 1.06) 1.16) 1.12)
1.42) 1.81) 1.73) 1.59) 1.90) 1.87) 1.24) 1.67) 1.31)
(0.72; (0.65; (0.58; (0.93; (0.84; (0.81;
(0.51; (0.44; (0.45; (0.89; (0.50; (0.89; (0.52; (0.57; (0.47;
1.19) 1.74) 1.18) 1.19) 1.36) 1.16)
0.85) 1.36) 1.06) 1.58) 1.61) 1.91) 0.83) 1.46) 0.93)
Adjusted for age, gender, comorbidity. Adjusted for age, gender, comorbidity, stage.
economic group with a depressive episode had the lowest high stage ratio (odds ratio ¼ 0.72, confidence interval 0.47; 1.09). The depression group had no effect on the duration of the primary investigation or on the duration from primary investigation to treatment after adjustment for age, gender, comorbidity and clinical stage. Cox regression analyses adjusted for age, gender, comorbidity and clinical stage showed that patients with depression had a significantly lower ratio for treatment (hazard ratio ¼ 0.79, confidence interval 0.64; 0.96). There was no difference in the ratio of receiving treatment for the patients with a depressive episode versus patients without depression. Patients with a recurrent depressive disorder had the lowest treatment ratio [hazard ratio ¼ 0.66, confidence interval 0.51; 0.85 for all patients, 0.78 (0.44; 1.36) for the low socio-economic group and 0.69 (0.45; 1.06) for the high socio-economic group]. There was no statistically significant effect of depression on the ratio of surgical treatment. However, patients with depression from a high socio-economic group had a higher ratio of surgical treatment [hazard ratio ¼ 1.25, confidence interval 0.93; 1.68 for depression versus none; 1.18 (0.75; 1.87) for a depressive episode versus none; 1.30 (0.89; 1.91) for a recurrent depressive disorder versus none], whereas patients with depression from a low socio-economic group had a slightly lower ratio for surgical treatment [hazard ratio ¼ 0.93, confidence interval 0.61; 1.43 for depression
versus none, 0.97 (0.53; 1.90) for a depressive episode versus none and 0.90 (0.50; 1.61) for a recurrent depressive disorder versus none]. Moreover, patients with depression had a significantly lower ratio for oncological treatment [hazard ratio ¼ 0.76, confidence interval 0.64; 0.91 for all patients, 0.95 (0.68; 1.35) for a low socio-economic group and 0.73 (0.57; 0.96) for a high socio-economic group]. There was no difference in the ratio of receiving oncological treatment for the patients with a depressive episode versus patients without depression. Patients with a recurrent depressive disorder had the lowest oncological treatment ratio [hazard ratio ¼ 0.65, confidence interval 0.52; 0.83 for all patients, 0.91 (0.57; 1.46) for a low socio-economic group and 0.66 (0.47; 0.93) for a high socio-economic group].
Discussion In our population of lung cancer patients 1.8% had at least one hospital contact with a diagnosis of depression during a 10 year period, which is very close to the expected according to the Danish Depression Database (about 2%). It thus seems reasonable to assume that our study population is representative of the Danish population regarding the number of patients with a depression diagnosis. There were twice as many women in our study population and that is also what you would expect to see in the entire Danish population.
Please cite this article in press as: Iachina M, et al., History of Depression in Lung Cancer Patients: Impact of Delay, Clinical Oncology (2017), http://dx.doi.org/10.1016/j.clon.2017.03.014
M. Iachina et al. / Clinical Oncology xxx (2017) 1e8
7
Staging Process
Strengths and Limitations
Both the unadjusted and adjusted analyses indicate that in a high socio-economic group there are fewer patients with a higher stage among patients with depression than among patients without depression. It is well known that patients from the high socio-economic group are more attentive to their health and more frequently see their general practitioner [14]. Patients with depression and especially a history of a recurrent depressive disorder, usually get a prophylactic medication to avoid a depression relapse. The use of such drugs requires regular monitoring, so patients who have depression in the anamnesis are more likely to see their general practitioner more often for control or treatment for depression. Patients from the high socio-economic group with a history of depression thus tend to go most frequently to their general practitioner and thereby stand a better chance of being referred to hospital for lung cancer treatment earlier in the course of the disease. Moreover, patients who are aware of their condition and have depression in the anamnesis will consult their general practitioner as soon as they notice depression symptoms. Some depression symptoms can resemble cancer symptoms and enable the general practitioner to send a patient for a cancer diagnosis earlier in the course of the disease. For all socio-economic groups as well as all depression groups we could not find any difference in either the time of the primary investigation or the time from the primary investigation to treatment, which indicates that depression has no effect on the duration of the staging process.
We used exclusively register data for this study. The strength of using the register data is the completeness of the study populations and independently collected data, which minimises the selection bias and makes it possible to look at sub-populations, e.g. persons with a special combination of socio-economic status. In the depression group we included all patients who had been diagnosed with depression within 10 years before the start of the lung cancer diagnostics. Some of them could have been cured of their psychiatric disorder long before they had a lung cancer diagnosis. On the other hand, we included in the depression group only in-patients and outpatients who had contact with a hospital with a depression diagnosis. There may have been some patients in our no depression group with a depression diagnosis who were treated by their general practitioner. Therefore, all estimated effects of depression could be underestimated. In the future it could be of interest to repeat the analyses and instead of using the Danish Depression Database, patients could be divided into depression and non-depression groups according to their use of anti-depression medicine. Most patients will react with depressive symptoms on receiving the cancer diagnosis, as getting such a heavy diagnosis is a powerful stress factor. A general characteristic of the patients with a history of depression is an increased sensitivity to stress factors, which with high probability can cause a new depression episode in this patient group and this is even more distinct in patients with a recurrent depression disorder in their history. For oncologists it could be difficult to distinguish between a normal patient reaction to a cancer diagnosis with depressive symptoms and a new depression episode. Depression affects the patient’s cognitive function and then difficult patients comply regarding oncological treatment. Therefore we think it will be expedient to involve a psychiatrist in the treatment process for patients with a history of depression.
Choice of Treatment For both socio-economic groups we found no effect of a depressive episode when receiving treatment. This means, with regards to receiving treatment, that patients with a depressive episode in the anamnesis are not different from the patients without depression in the anamnesis. Our results suggest that the patients with a recurrent depressive disorder from a high socio-economic group have a higher chance of receiving surgical treatment even after adjusting for age, gender, stage and comorbidity. As mentioned before, patients from the high socio-economic group with a history of depression will more often consult their general practitioner. They will probably have comorbidities, such as hypertension and diabetes etc., under control and thus have a high chance of receiving surgical treatment. We found that patients with a recurrent depressive disorder from both socio-economic groups had a lower chance of receiving treatment in general. Moreover, the patients with a recurrent depressive disorder from a high socioeconomic group had a lower chance of receiving oncological treatment. The reason for this could be that a patient with a recurrent depressive disorder will probably develop a new depression episode after the cancer diagnosis and then choose not to receive palliative care.
Conclusion The present study has shown that a depressive episode in a patient’s anamnesis has no effect on the delay in diagnosis and the start of treatment for lung cancer. However, a history of a recurrent depressive disorder leads to a reduced probability of receiving treatment for lung cancer. Therefore, these patients need special attention when diagnosed with lung cancer.
References [1] Sogaard M, Thomsen RW, Bossen KS, Sorensen HT, Norgaard M. The impact of comorbidity on cancer survival: a review. Clin Epidemiol 2013;5(Suppl. 1):3e29. [2] Iachina M, Green A, Jakobsen E. The direct and indirect impact of comorbidity on the survival of patients with non-small cell lung cancer: a combination of survival, staging and resection models with missing measurements in covariates. BMJ Open 2014;4(2). e003846.
Please cite this article in press as: Iachina M, et al., History of Depression in Lung Cancer Patients: Impact of Delay, Clinical Oncology (2017), http://dx.doi.org/10.1016/j.clon.2017.03.014
8
M. Iachina et al. / Clinical Oncology xxx (2017) 1e8
[3] Pinquart M, Duberstein PR. Depression and cancer mortality: a meta-analysis. Psychol Med 2010;40(11):1797e1810. [4] Sullivan DR, Ganzini L, Duckart JP, et al. Treatment receipt and outcomes among lung cancer patients with depression. Clin Oncol 2014;26(1):25e31. [5] Spiegel D, Giese-Davis J. Depression and cancer: mechanisms and disease progression. Biol Psychiatr 2003;54(3):269e282. [6] DiMatteo MR, Lepper HS, Croghan TW. Depression is a risk factor for noncompliance with medical treatment: metaanalysis of the effects of anxiety and depression on patient adherence. Arch Intern Med 2000;160(14):2101e2107. [7] Lasser K, Wesley JB, Woolhandler S, et al. Smoking and mental illness: a population based prevalence study. Am Med Assoc 2000;284:2606e2610. [8] Murata C, Kondo K, Hirai H, et al. Association between depression and socio-economic status among communitydwelling elderly in Japan: the Aichi Gerontological Evaluation Study (AGES). Health Place 2008;14(3):406e414. [9] Butterworth P, Rodgers B, Windsor TD. Financial hardship, socio-economic position and depression: results from the PATH Through Life Survey. Soc Sci Med 2009;69:229e237.
[10] Skapinakis P, Weich S, Lewis G, et al. Socio-economic position and common mental disorders: longitudinal study in the general population in the UK. Br J Psychiatr 2006;198:109e117. [11] Jakobsen E, Green A, Oesterlind K, Rasmussen TR, Iachina M, Palshof T. Nationwide quality improvement in lung cancer care: the role of the Danish Lung Cancer Group and Registry. J Thoracic Oncol 2013;8(10):1238e1247. [12] Lynge E, Sandegaard JL, Rebolj M. The Danish National Patient Register. Scand J Public Health 2011;39(Suppl. 7):S30eS33. [13] http://www.kcks-vest.dk/kliniskeþkvalitetsdatabaser/ danskþdepressionsdatabase. [14] Charlson ME, Pompei P, Ales KL, et al. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis 1987;40:373e383. [15] Thygesen LC, Daasnes C, Thaulow I, Hansen HB. Introduction to Danish (nationwide) registers on health and social issues: structure, access, legislation and archiving. Scand J Public Health 2011;39:12e16. [16] Pedersen CB. The Danish Civil Registration System. Scand J Public Health 2011;39:22e25. http://dx.doi.org/10.1177/ 1403494810387965.
Please cite this article in press as: Iachina M, et al., History of Depression in Lung Cancer Patients: Impact of Delay, Clinical Oncology (2017), http://dx.doi.org/10.1016/j.clon.2017.03.014
Clinical Oncology journal homepage: www.clinicaloncologyonline.net
Original Article
History of Depression in Lung Cancer Patients: Impact of Delay M. Iachina *y, M.M. Brønserud zx, E. Jakobsen zx, O. Trosko {, A. Green x * Center
for Clinical Epidemiology, Odense University Hospital, Odense, Denmark The Institute of Clinical Research, University of Southern Denmark, Odense, Denmark z Department of Thoracic Surgery, Odense University Hospital, Odense, Denmark x OPEN, Odense Patient Data Exploratory Network, Odense University Hospital/Department of Clinical Research, University of Southern Denmark, Odense, Denmark { Research Unit of Psychiatry, Department of Clinical Research, University of Southern Denmark, Odense, Denmark y
Received 6 September 2016; received in revised form 20 March 2017; accepted 21 March 2017
Abstract Aims: To examine the influence of a history of depression in the process of diagnostic evaluation and the choice of treatment in lung cancer. Materials and methods: The analysis was based on all patients with non-small cell lung cancer who were registered in 2008e2014; in total, 27 234 patients. To estimate the effect of depression on the diagnostic process and the choice of treatment in lung cancer we fitted a logistic regression model and a Cox regression model adjusting for age, gender, resection and stage. Results: Depression in a patient’s anamnesis had no significant effect on the delay in diagnostic evaluation (hazard ratio ¼ 0.99 with 95% confidence interval 0.90; 1.09). Patients with a history of periodic depression had a 33% lower treatment rate (odds ratio ¼ 0.66 with 95% confidence interval 0.51; 0.85) than patients without a history of depression. Conclusions: Our study shows that patients with a history of periodic depression need special attention when diagnosed with lung cancer. Ó 2017 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved.
Key words: Depression; diagnostic evaluation; lung cancer; socio-economic status; treatment
Introduction With an annual incidence of 4500 patients, primary lung cancer is one of the most common cancers in Denmark. In general, the lung cancer prognosis is poor, with a 5 year survival of about 10e12%. Comorbidity is common in patients with lung cancer (26e81%) [1] and an increased comorbidity increases mortality significantly [2]. Depression has been associated with increased mortality in patients with lung cancer [3,4], but not much is known as to the reason. Spiegel and Giese-Davis [5] identified several reasons why depression may have an effect on the mortality of cancer patients in general. First, depression may have a pathophysiological effect via a neuroendocrine and
Author for correspondence: M. Iachina, Center for Clinical Epidemiology, Odense University Hospital, Sdr. Boulevard 29, Entrance 216, Ground Floor East, DK-5000 Odense C, Denmark. Tel: þ45-21158110. E-mail address: [email protected] (M. Iachina).
immunological function that influences mortality. Second, many of the symptoms of cancer and the side-effects of its treatment are similar to those of depression, e.g. sleep disturbance, anorexia, fatigue and concentration difficulties. Therefore, referral to an examination for cancer can be delayed for depressed patients. Third, patients with depression may be less likely to adhere to preventive screening procedures, cancer treatments or recommendations for maintaining good health. Depression and smoking are associated [6] and lung cancer-specific symptoms, such as cough and shortness of breath, are common in heavy smokers, thus hiding the lung cancer symptoms and adding further delay to the diagnosis. Depressed patients may miss screening appointments and are thus more likely to delay the treatment and less likely to receive treatment or to implement treatment [7]. Diagnostic procedures in suspected lung cancer are carried out to establish a diagnosis and the stage of lung cancer. The choice of treatment is furthermore based on the patient
http://dx.doi.org/10.1016/j.clon.2017.03.014 0936-6555/Ó 2017 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved.
Please cite this article in press as: Iachina M, et al., History of Depression in Lung Cancer Patients: Impact of Delay, Clinical Oncology (2017), http://dx.doi.org/10.1016/j.clon.2017.03.014
2
M. Iachina et al. / Clinical Oncology xxx (2017) 1e8
performance and comorbidity. Patients may be treated with surgical resection or chemo- and/or radiotherapy. Recently targeted and immunotherapy are added to treatment options. Surgical resection of the tumour is associated with the best survival, but only about 20% of the patients are eligible for resection. The clinical stage and performance are the most important factors when deciding on the treatment. A range of other prognostic factors are also taken into account before the final decision about treatment is made, including age, smoking, alcohol or drug abuse, patient preferences and comorbidity. The aim of this study was to examine the influence of the history of depression in the process of diagnostic evaluation and the choice of treatment in lung cancer. A similar study has previously been carried out in the USA by Sullivan et al. [4], but without taking into account the differences in the patient’s socio-economic status. They found no association of depression diagnosis with receiving treatment. It is well known that there is a strong association between socioeconomic position and the experience of depression [8] and that a low socio-economic status is associated with late-stage cancer diagnoses and the type of treatment received [9,10]. Here we investigated the effect of a history of depression when adjusting for the patient’s socioeconomic status.
Materials and Methods We used data from the four national registries: The Danish Lung Cancer Registry We used the Danish Lung Cancer Registry (DLCR) to identify patients in Denmark with lung cancer. Since 2000 the DLCR has monitored and evaluated the quality of treatment effort concerning lung cancer [11]. All Danish lung cancer patients are included in the DLCR and to this date the database contains data on more than 55 000 cases of lung cancer. Between 2000 and 2002 clinicians identified and reported new patients to the DLCR, but since 2003 lung cancer patients have been identified in the Danish National Patient Registry (DNPR), where the first occurrence of the diagnostic codes DC34 and DC33, according to the International Classification of Diseases 10 (ICD-10), are identified. Information about these patients and their activities (procedures and treatments) registered with the DNPR, together with information from the Danish Pathology Registry, provides the basis of establishing patient trajectories and associated diagnostic procedures and treatments in the DLCR. Only about 75% of the information required for the DLCR is available in the Danish central registries and the participating clinicians are requested to supplement with the remaining about 25% of information from their medical records. As participation in this supplementation is mandatory by law, data completeness is very high (more than 95%). Data on comorbidity were obtained from the DNPR [12]. The history and status on depression is not
requested routinely by the DLCR and was obtained specifically for this project. The Danish Depression Database We retrieved data from the Danish Depression Database [13] to identify patients with a history of depression. The database includes all patients who are 18 years or older with a diagnosis of depression, have permanent residence in Denmark and are hospitalised or affiliated with a psychiatric hospital in Denmark. The disease is diagnosed according to ICD-10 criteria and includes all sub-codes: DF32.X - depressive episode (DF32.0; DF32.1; DF32.2; DF32.3; DF32.8; DF32.9); DF33.X e recurrent depressive disorder (DF33.0; DF33.1; DF33.2; DF33.3; DF33.4; DF32.8; DF32.9); DF34.1X e dysthymia; DF06.32 e organic affective disorder, depressive. We included both in-patients and out-patients who had at least one hospital contact with depression (DF32.X depressive episode and DF33.X recurrent depressive disorder) within 10 years before the start of the lung cancer diagnostics process. We excluded from our analysis patients with dysthymia and organic depression as the definition and diagnostic delineation of these two groups is more uncertain. Here we will examine the effect of the following three groups of patients: one includes patients with the diagnosis ‘depressive episode’ (ICD-10 DF32.X), another includes patients with the diagnosis ‘recurrent depressive disorder’ (ICD-10 DF33.X) and the last one includes patients from both groups. The National Patient Register We included information on comorbidity for each patient up to 10 years before the lung cancer diagnosis, using the DNPR, which was established in 1977. This register contains data on all interventions related to diagnostic evaluation and treatment for somatic patient admissions in Denmark [14]. We used a modification of the Charlson comorbidity index (CCI) for the classification of somatic comorbidity [15]. All hospital contacts represented with a cancer diagnosis, registered within 150 days before the date of lung cancer diagnosis, were excluded from contribution to the cancer comorbidity group in order to avoid the influence of misclassification of cases with lung cancer before the final diagnosis. The Income Statistics Register We included information on the socio-economic status of each patient at the time of the lung cancer diagnosis, using information available with The Income Statistics Register [16]. The socio-economic class was defined as household income in the year the patient was diagnosed with lung cancer, adjusted for the number of people in the household. The socio-economic class was divided by the median into two groups, low and high.
Please cite this article in press as: Iachina M, et al., History of Depression in Lung Cancer Patients: Impact of Delay, Clinical Oncology (2017), http://dx.doi.org/10.1016/j.clon.2017.03.014
M. Iachina et al. / Clinical Oncology xxx (2017) 1e8
All permanent residents in Denmark are assigned to a unique personal identification number, which is used in all Danish registers enabling data linkage [15]. Study Population The analysis is based on all patients with non-small cell lung cancer who were registered in 2009e2011, in total 27 234 patients. For each patient, comorbidity information was collected using the DNPR up to 10 years before the lung cancer diagnosis. This register contains data including coding of all interventions related to diagnostic evaluation and treatment for all somatic patient admissions in Denmark. Statistical Analyses For the sake of simplicity, we will in the following refer to the group of patients, who have had depression in the anamneses, as a group of patients with depression and correspondingly the group of patients who do not have depression in the anamneses will be called patients without depression. We used Fisher’s exact test to compare baseline characteristics of patients with and without depression, and to compare patients with a depressive episode to patients without depression, and patients with a recurrent depressive disorder to patients without depression. We used a logistic regression model, with clinical stage as outcome, to see if there was any delay in the referral to examination for lung cancer for depressed patients compared with non-depressed patients. A Cox regression model was used to estimate the effect of depression on the length of the primary investigations, and to estimate the effect of depression on the time from the primary investigation to the start to treatment, for those patients who received specific treatment. Using a logistic regression model we also estimated the effect of depression on the risk of receiving a treatment in general, receiving surgical treatment and receiving oncological treatment. The following variables acted as confounders in the prediction model: age: young age was defined as age lower than the mean age for the patient population; gender; comorbidity: 0, CCI score of 0; 1, CCI score of 1; 2, than CCI score >1; clinical tumour stage when possible. As there is an interaction between socio-economic status and the experience of depression, the socio-economic class was used as a stratification variable. The socio-economic class was taken from the patient’s personal income in the year of lung cancer diagnosis. The personal income for our study population ranged from 95 000 DKK to 668 000 DKK per year with a 25% fractile of 162 000 DKK, a 50% fractile of 200 000 DKK and a 75% fractile of 270 000 DKK. The variable socio-economic class was divided into: low e socioeconomic class being lower than the mean socioeconomic class for the patient population; high e socioeconomic class being equal or higher than the mean socio-economic class for the patient population. For all outcomes we carried out three separate analyses: first to compare patients with depression with
3
patients without depression; second to compare patients with a depressive episode (ICD-10 DF32) with patients without depression; third to compare patients with a recurrent depressive disorder (ICD-10 DF33) with patients without depression. Moreover, as an association between socio-economic status and depression is not unequivocal we also completed analyses stratified by socio-economic class. Ethical Aspects DLCR is approved as a national clinical database by the Danish National Board of Health (j.nr. 7-201-03-15/1). Permission for the present study was granted by the Danish Data Protection Agency (j.nr. 2008-58-0035).
Results In total we extracted 27 234 patients from the DLCR who had no hospital contact with diagnosed depression within 10 years before the primary lung cancer diagnosis and 508 who had hospital contact (see Table 1); 215 of those had a depressive episode, 293 had a recurrent depressive disorder. Patients with depression were younger and with a preponderance of females compared with patients without depression. Similar patterns of age and gender distribution were observed in patients with a depressive episode compared with patients with a recurrent depressive disorder. More patients from the depression group had a high comorbidity (42.5%) compared with the group of patients without depression (31.5%). Among patients with a depressive episode, 39.1% had a high comorbidity, whereas in patients with a recurrent depressive disorder 45.1% had a high comorbidity. There was no difference in the stage and socio-economic distribution for patients with depression regardless of depression group. There was no difference in the duration of the diagnostic process between all patient groups. Table 2 shows estimates of the univariate analyses. There was no difference in the stage distribution between patients with depression (53.9% patients with high stage) and without depression (57.3% patients with high stage, P ¼ 0.312). The same applies to the patients from the depressive episode group (53.5% patients with high stage) as well as to the patients from the recurrent depressive disorder group (54.3% patients with high stage). The results suggest that in the high socio-economic group patients with depression had a lower chance (51.0%) of having a high stage compared with the patients without depression (58.2%, P ¼ 0.037). There was no difference in the duration of the diagnostic evaluation process or in the days from the primary investigation to treatment among depression groups. The ratio for treatment was higher for patients without depression (75.4%) than for patients with depression (72.1%); 77.2% of the patients with a depressive episode received treatment, whereas the treatment ratio was only
Please cite this article in press as: Iachina M, et al., History of Depression in Lung Cancer Patients: Impact of Delay, Clinical Oncology (2017), http://dx.doi.org/10.1016/j.clon.2017.03.014
0.106 0.753
72 (24.6%) 133 (45.4%) 88 (30.0%)
0.312
0.227
8714 (32.6%) 15 310 (57.3%) 2702 (10.1%)
8714 (30.2%) 11 313 (42.3%) 7332 (27.4%)
64 (29.8%) 96 (44.7%) 55 (25.6%)
0.576 0.527
103 (35.1%) 159 (54.3%) 31 (10.6%)
<0.000
77 (35.8%) 115 (53.5%) 23 (10.7%)
<0.000 0.053
105 (35.8%) 56 (19.1%) 132 (45.1%) 12 304 (46.0%) 5998 (22.4%) 8424 (31.5%)
91 (42.3%) 40 (18.6%) 84 (39.1%)
<0.000 <0.000 293 168 (57.3%) 97 (33.1%) <0.000 0.027 215 125 (58.1%) 95 (44.2%) <0.000 <0.000 26 726 12 231 (45.8%) 13 838 (51.8%)
Number 508 Age (young) 293 (57.7%) Gender (male) 192 (37.8%) Comorbidity 0 196 (38.6%) 1 96 (18.9%) >1 216 (42.5%) Stage 0, I, II, IIIa 180 (35.4%) IIIb, IV 274 (53.9%) Missing 54 (10.6%) Socio-economic status Low 136 (26.8%) High 229 (45.1%) Missing 143 (28.2%)
With depression
Without depression
P value for test depression versus none
Depression episode
P value for test depression episode versus none
Recurrent depressive disorder
P value for test recurrent depressive disorder versus none
M. Iachina et al. / Clinical Oncology xxx (2017) 1e8 Table 1 Baseline characteristics of patients with depression and without depression, as well as of patients with a single depression episode and patients with a history of periodic depression
4
68.3% for the patients with a recurrent depressive disorder. The difference in the treatment ratio between patients with a recurrent depressive disorder and patients without depression was statistically significant (P ¼ 0.005). Only 79.1% of the patients from the high socio-economic group with recurrent depressive disorder received treatment against 84.9% of the patients from the high socio-economic group without depression. There was almost no difference in the ratio for surgical treatment in the patients from the low socio-economic groups (20.1% for the patients without depression, 19.9% for the patients with depression, 20.3% for the patients with a depressive episode and 19.4% for the patients with a recurrent depressive disorder). Patients from the high socio-economic group had a substantially higher ratio for surgical treatment (22.5% for the patients without depression, 27.5% for the patients with depression, 26.0% for the patients with a depressive episode and 28.6% for the patients with a recurrent depressive disorder). The patients with depression received statistically significantly less oncological treatment than the patients without depression (58.5% for the patients without depression versus 51.9% for the patients with depression). There was no difference in the ratio for oncological treatment between the group of patients with a depressive episode (57.7%) and the group of patients without depression (58.5%), but only 47.8% of the patients with the recurrent depressive disorder received oncological treatment. The difference in the oncological treatment ratio between patients with a recurrent depressive disorder and patients without depression was statistically significant (P < 0.001). The difference in the ratio for oncological treatment for the patients from the low socio-economic groups was not significant (59.5% for the patients without depression, 58.1% for the patients with depression, 59.4% for the patients with a depressive episode and 56.9% for the patients with a recurrent depressive disorder). The ratio for oncological treatment was significantly lower (P ¼ 0.015) for patients with depression from the high socio-economic group (54.6%) compared with patients without depression (62.3%). The patients with a recurrent depressive disorder from the high socio-economic group had the lowest ratio for oncological treatment (51.1%) among the patients from the high socio-economic group. Table 3 shows estimates of the multivariate analyses. In each model we used age, gender, comorbidity as a confounder and stage where possible. The adjusted results for the stage confirmed the unadjusted findings shown in Table 2. There was no difference in the stage distribution between patients with depression and without depression (odds ratio ¼ 0.9, confidence interval 0.74; 1.09). The same applies to the patients from the depressive episode group (odds ratio ¼ 0.86, confidence interval 0.65; 1.16) as well as for the patients from the recurrent depressive disorder group (odds ratio ¼ 0.93, 0.72; 1.19). The results indicate that the ratio for a high stage was lower for patients with depression in the high socio-economic group (odds ratio ¼ 0.78, confidence interval 0.59; 1.03) and patients from the high socio-
Please cite this article in press as: Iachina M, et al., History of Depression in Lung Cancer Patients: Impact of Delay, Clinical Oncology (2017), http://dx.doi.org/10.1016/j.clon.2017.03.014
Outcome
Socio-economic class
With depression
Without depression
P value for test depression versus none
Depression episode
P value for test depression episode versus none
Recurrent depressive disorder
P value for test recurrent depressive disorder versus none
Stage (high versus low)
All Low High All Low High
274 (53.9%) 67 (54.5%) 106 (51.0%) 30.1 (3.4) 33.6 (6.9) 35.6 (6.2)
15 310 (57.3%) 4066 (55.5%) 6047 (58.2%) 28.8 (0.5) 31.9 (0.9) 32.8 (0.9)
0.312 0.903 0.037 0.731 0.829 0.668
115 (53.5%) 30 (53.6%) 44 (50%) 25.8 (1.9) 25.9 (2.0) 28.6 (3.6)
0.527 0.779 0.122 0.594 0.592 0.672
159 (54.3) 37 (56.1%) 62 (51.7%) 33.3 (5.7) 40.3 (12.8) 40.7 (10.6)
0.576 0.922 0.155 0.369 0.421 0.3646
All Low High
42.3 (2.1) 42.3 (2.1) 43.4 (40.9)
40.7 (0.4) 44.9 (0.8) 40.9 (0.52)
0.551 0.679 0.509
44.6 (4.2) 45.3 (2.9) 46.3 (7.8)
0.344 0.969 0.343
40.5 (1.6) 39.6 (3.0) 41.1 (1.9)
0.996 0.543 0.997
All Low High All Low High All Low High
366 (72.1%) 106 (77.9%) 188 (82.1%) 102 (20.1%) 27 (19.9%) 63 (27.5%) 264 (51.9%) 79 (58.1%) 125 (54.6%)
20 144 (75.4%) 6436 (79.6%) 9614 (84.9%) 4500 (16.8%) 1625 (20.1%) 2546 (22.5%) 15 644 (58.5%) 4811 (59.5%) 7068 (62.25%)
0.085 0.239 0.277 0.053 0.941 0.073 0.003 0.733 0.015
166 (77.2%) 51 (79.7%) 82 (85.4%) 42 (19.5%) 13 (20.3%) 25 (26,0%) 124 (57.7%) 38 (59.4%) 57 (59.4%)
0.388 0.998 0.901 0.290 0.969 0.406 0.795 0.976 0.533
200 (68.3%) 55 (76.4%) 106 (79.1%) 60 (20.5%) 14 (19.4%) 38 (28.6%) 149 (47.8%) 41 (56.9%) 68 (51.1%)
0.005 0.490 0.091 0.101 0.886 0.100 0.000 0.653 0.008
Days of primary investigation Mean (standard deviation) Days from primary investigation to treatment Mean (standard deviation) Treatment (yes versus no) Resection (yes versus no) Oncological treatment (yes versus no)
M. Iachina et al. / Clinical Oncology xxx (2017) 1e8 5
Please cite this article in press as: Iachina M, et al., History of Depression in Lung Cancer Patients: Impact of Delay, Clinical Oncology (2017), http://dx.doi.org/10.1016/j.clon.2017.03.014
Table 2 Effect of depression (yes versus no) for different socio-economic patient groups on staging process and treatment
6
M. Iachina et al. / Clinical Oncology xxx (2017) 1e8
Table 3 Effect of depression (yes versus no) estimated in the models that describe the primary investigation and the treatment offers; logistic regressions are shown as odds ratio (95% confidence interval), Cox regressions are shown as hazard ratio (95% confidence interval) Outcome
Socio-economic class
Depression versus none
Depression episode versus none
Recurrent depressive disorder versus none
Stage (odds ratio)* (high versus low)
All Low High All Low High
0.90 1.01 0.78 0.99 0.98 0.94
0.86 0.96 0.72 0.92 0.89 0.91
0.93 1.07 0.83 1.05 1.06 0.97
All Low High
0.91 (0.81; 1.01) 0.96 (0.78; 1,17) 0.87 (0.75; 1.01)
0.88 (0.75; 1.03) 0.85 (0.63; 1.13) 0.86 (0.69; 1.08)
0.93 (0.80; 1.07) 1.07 (0.82; 1.41) 0.88 (0.72; 1.07)
All Low High All Low High All Low High
0.79 0.86 0.79 1.16 0.93 1.25 0.76 0.95 0.73
1.02 0.97 0.97 1.13 0.97 1.18 0.94 1.01 0.87
0.66 0.78 0.69 1.18 0.90 1.30 0.65 0.91 0.66
Days of primary investigation (hazard ratio)y Mean (standard deviation) Days from primary investigation to treatment (hazard ratio)y Mean (standard deviation) Treatment (odds ratio)y (yes versus no) Resection (odds ratio)y (yes versus no) Oncologic treatment (odds ratio)y (yes versus no) * y
(0.74; (0.71; (0.59; (0.90; (0.82; (0.82;
(0.64; (0.57; (0.55; (0.93; (0.61; (0.93; (0.64; (0.68; (0.57;
1.09) 1.46) 1.03) 1.09) 1.17) 1.08)
0.96) 1.31) 1.11) 1.45) 1.43) 1.68) 0.91) 1.35) 0.96)
(0.65; (0.56; (0.47; (0.80; (0.68; (0.74;
(0.74; (0.52; (0.55; (0.81; (0.53; (0.75; (0.72; (0.61; (0.57;
1.16) 1.62) 1.09) 1.06) 1.16) 1.12)
1.42) 1.81) 1.73) 1.59) 1.90) 1.87) 1.24) 1.67) 1.31)
(0.72; (0.65; (0.58; (0.93; (0.84; (0.81;
(0.51; (0.44; (0.45; (0.89; (0.50; (0.89; (0.52; (0.57; (0.47;
1.19) 1.74) 1.18) 1.19) 1.36) 1.16)
0.85) 1.36) 1.06) 1.58) 1.61) 1.91) 0.83) 1.46) 0.93)
Adjusted for age, gender, comorbidity. Adjusted for age, gender, comorbidity, stage.
economic group with a depressive episode had the lowest high stage ratio (odds ratio ¼ 0.72, confidence interval 0.47; 1.09). The depression group had no effect on the duration of the primary investigation or on the duration from primary investigation to treatment after adjustment for age, gender, comorbidity and clinical stage. Cox regression analyses adjusted for age, gender, comorbidity and clinical stage showed that patients with depression had a significantly lower ratio for treatment (hazard ratio ¼ 0.79, confidence interval 0.64; 0.96). There was no difference in the ratio of receiving treatment for the patients with a depressive episode versus patients without depression. Patients with a recurrent depressive disorder had the lowest treatment ratio [hazard ratio ¼ 0.66, confidence interval 0.51; 0.85 for all patients, 0.78 (0.44; 1.36) for the low socio-economic group and 0.69 (0.45; 1.06) for the high socio-economic group]. There was no statistically significant effect of depression on the ratio of surgical treatment. However, patients with depression from a high socio-economic group had a higher ratio of surgical treatment [hazard ratio ¼ 1.25, confidence interval 0.93; 1.68 for depression versus none; 1.18 (0.75; 1.87) for a depressive episode versus none; 1.30 (0.89; 1.91) for a recurrent depressive disorder versus none], whereas patients with depression from a low socio-economic group had a slightly lower ratio for surgical treatment [hazard ratio ¼ 0.93, confidence interval 0.61; 1.43 for depression
versus none, 0.97 (0.53; 1.90) for a depressive episode versus none and 0.90 (0.50; 1.61) for a recurrent depressive disorder versus none]. Moreover, patients with depression had a significantly lower ratio for oncological treatment [hazard ratio ¼ 0.76, confidence interval 0.64; 0.91 for all patients, 0.95 (0.68; 1.35) for a low socio-economic group and 0.73 (0.57; 0.96) for a high socio-economic group]. There was no difference in the ratio of receiving oncological treatment for the patients with a depressive episode versus patients without depression. Patients with a recurrent depressive disorder had the lowest oncological treatment ratio [hazard ratio ¼ 0.65, confidence interval 0.52; 0.83 for all patients, 0.91 (0.57; 1.46) for a low socio-economic group and 0.66 (0.47; 0.93) for a high socio-economic group].
Discussion In our population of lung cancer patients 1.8% had at least one hospital contact with a diagnosis of depression during a 10 year period, which is very close to the expected according to the Danish Depression Database (about 2%). It thus seems reasonable to assume that our study population is representative of the Danish population regarding the number of patients with a depression diagnosis. There were twice as many women in our study population and that is also what you would expect to see in the entire Danish population.
Please cite this article in press as: Iachina M, et al., History of Depression in Lung Cancer Patients: Impact of Delay, Clinical Oncology (2017), http://dx.doi.org/10.1016/j.clon.2017.03.014
M. Iachina et al. / Clinical Oncology xxx (2017) 1e8
7
Staging Process
Strengths and Limitations
Both the unadjusted and adjusted analyses indicate that in a high socio-economic group there are fewer patients with a higher stage among patients with depression than among patients without depression. It is well known that patients from the high socio-economic group are more attentive to their health and more frequently see their general practitioner [14]. Patients with depression and especially a history of a recurrent depressive disorder, usually get a prophylactic medication to avoid a depression relapse. The use of such drugs requires regular monitoring, so patients who have depression in the anamnesis are more likely to see their general practitioner more often for control or treatment for depression. Patients from the high socio-economic group with a history of depression thus tend to go most frequently to their general practitioner and thereby stand a better chance of being referred to hospital for lung cancer treatment earlier in the course of the disease. Moreover, patients who are aware of their condition and have depression in the anamnesis will consult their general practitioner as soon as they notice depression symptoms. Some depression symptoms can resemble cancer symptoms and enable the general practitioner to send a patient for a cancer diagnosis earlier in the course of the disease. For all socio-economic groups as well as all depression groups we could not find any difference in either the time of the primary investigation or the time from the primary investigation to treatment, which indicates that depression has no effect on the duration of the staging process.
We used exclusively register data for this study. The strength of using the register data is the completeness of the study populations and independently collected data, which minimises the selection bias and makes it possible to look at sub-populations, e.g. persons with a special combination of socio-economic status. In the depression group we included all patients who had been diagnosed with depression within 10 years before the start of the lung cancer diagnostics. Some of them could have been cured of their psychiatric disorder long before they had a lung cancer diagnosis. On the other hand, we included in the depression group only in-patients and outpatients who had contact with a hospital with a depression diagnosis. There may have been some patients in our no depression group with a depression diagnosis who were treated by their general practitioner. Therefore, all estimated effects of depression could be underestimated. In the future it could be of interest to repeat the analyses and instead of using the Danish Depression Database, patients could be divided into depression and non-depression groups according to their use of anti-depression medicine. Most patients will react with depressive symptoms on receiving the cancer diagnosis, as getting such a heavy diagnosis is a powerful stress factor. A general characteristic of the patients with a history of depression is an increased sensitivity to stress factors, which with high probability can cause a new depression episode in this patient group and this is even more distinct in patients with a recurrent depression disorder in their history. For oncologists it could be difficult to distinguish between a normal patient reaction to a cancer diagnosis with depressive symptoms and a new depression episode. Depression affects the patient’s cognitive function and then difficult patients comply regarding oncological treatment. Therefore we think it will be expedient to involve a psychiatrist in the treatment process for patients with a history of depression.
Choice of Treatment For both socio-economic groups we found no effect of a depressive episode when receiving treatment. This means, with regards to receiving treatment, that patients with a depressive episode in the anamnesis are not different from the patients without depression in the anamnesis. Our results suggest that the patients with a recurrent depressive disorder from a high socio-economic group have a higher chance of receiving surgical treatment even after adjusting for age, gender, stage and comorbidity. As mentioned before, patients from the high socio-economic group with a history of depression will more often consult their general practitioner. They will probably have comorbidities, such as hypertension and diabetes etc., under control and thus have a high chance of receiving surgical treatment. We found that patients with a recurrent depressive disorder from both socio-economic groups had a lower chance of receiving treatment in general. Moreover, the patients with a recurrent depressive disorder from a high socioeconomic group had a lower chance of receiving oncological treatment. The reason for this could be that a patient with a recurrent depressive disorder will probably develop a new depression episode after the cancer diagnosis and then choose not to receive palliative care.
Conclusion The present study has shown that a depressive episode in a patient’s anamnesis has no effect on the delay in diagnosis and the start of treatment for lung cancer. However, a history of a recurrent depressive disorder leads to a reduced probability of receiving treatment for lung cancer. Therefore, these patients need special attention when diagnosed with lung cancer.
References [1] Sogaard M, Thomsen RW, Bossen KS, Sorensen HT, Norgaard M. The impact of comorbidity on cancer survival: a review. Clin Epidemiol 2013;5(Suppl. 1):3e29. [2] Iachina M, Green A, Jakobsen E. The direct and indirect impact of comorbidity on the survival of patients with non-small cell lung cancer: a combination of survival, staging and resection models with missing measurements in covariates. BMJ Open 2014;4(2). e003846.
Please cite this article in press as: Iachina M, et al., History of Depression in Lung Cancer Patients: Impact of Delay, Clinical Oncology (2017), http://dx.doi.org/10.1016/j.clon.2017.03.014
8
M. Iachina et al. / Clinical Oncology xxx (2017) 1e8
[3] Pinquart M, Duberstein PR. Depression and cancer mortality: a meta-analysis. Psychol Med 2010;40(11):1797e1810. [4] Sullivan DR, Ganzini L, Duckart JP, et al. Treatment receipt and outcomes among lung cancer patients with depression. Clin Oncol 2014;26(1):25e31. [5] Spiegel D, Giese-Davis J. Depression and cancer: mechanisms and disease progression. Biol Psychiatr 2003;54(3):269e282. [6] DiMatteo MR, Lepper HS, Croghan TW. Depression is a risk factor for noncompliance with medical treatment: metaanalysis of the effects of anxiety and depression on patient adherence. Arch Intern Med 2000;160(14):2101e2107. [7] Lasser K, Wesley JB, Woolhandler S, et al. Smoking and mental illness: a population based prevalence study. Am Med Assoc 2000;284:2606e2610. [8] Murata C, Kondo K, Hirai H, et al. Association between depression and socio-economic status among communitydwelling elderly in Japan: the Aichi Gerontological Evaluation Study (AGES). Health Place 2008;14(3):406e414. [9] Butterworth P, Rodgers B, Windsor TD. Financial hardship, socio-economic position and depression: results from the PATH Through Life Survey. Soc Sci Med 2009;69:229e237.
[10] Skapinakis P, Weich S, Lewis G, et al. Socio-economic position and common mental disorders: longitudinal study in the general population in the UK. Br J Psychiatr 2006;198:109e117. [11] Jakobsen E, Green A, Oesterlind K, Rasmussen TR, Iachina M, Palshof T. Nationwide quality improvement in lung cancer care: the role of the Danish Lung Cancer Group and Registry. J Thoracic Oncol 2013;8(10):1238e1247. [12] Lynge E, Sandegaard JL, Rebolj M. The Danish National Patient Register. Scand J Public Health 2011;39(Suppl. 7):S30eS33. [13] http://www.kcks-vest.dk/kliniskeþkvalitetsdatabaser/ danskþdepressionsdatabase. [14] Charlson ME, Pompei P, Ales KL, et al. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis 1987;40:373e383. [15] Thygesen LC, Daasnes C, Thaulow I, Hansen HB. Introduction to Danish (nationwide) registers on health and social issues: structure, access, legislation and archiving. Scand J Public Health 2011;39:12e16. [16] Pedersen CB. The Danish Civil Registration System. Scand J Public Health 2011;39:22e25. http://dx.doi.org/10.1177/ 1403494810387965.
Please cite this article in press as: Iachina M, et al., History of Depression in Lung Cancer Patients: Impact of Delay, Clinical Oncology (2017), http://dx.doi.org/10.1016/j.clon.2017.03.014