High level of distress in long-term survivors of thyroid carcinoma: Results of rapid screening using the distress thermometer

Abstract

Context. Cancer patients are at increased risk for distress. The Distress Thermometer (DT) and problem list (PL) are short-tools validated and recommended for distress screening in cancer patients. Objective. To investigate the level of distress and problems experienced by survivors of differentiated non-medullary thyroid carcinoma (DTC), using the DT and PL and whether this correlates with clinical and demographical variables. Participants, design and setting. All 205 DTC patients, under follow-up at the outpatient clinic of our university hospital, were asked to fill in the DT and PL, hospital anxiety and depression scale (HADS), illness cognition questionnaire (ICQ) and an ad hoc questionnaire. Receiver Operator Characteristic analysis (ROC) was used to establish the optimal DT cut-off score according to HADS. Correlations of questionnaires scores with data on diagnosis, treatment and follow-up collected from medical records were analyzed. Results. Of the 159 respondents, 145 agreed to participate [118 in remission, median follow-up 7.2 years (range 3 months–41 years)]. Of these, 34.3% rated their distress score ≥5, indicating clinically relevant distress according to ROC analysis. Patients reported physical (86%) over emotional problems (76%) as sources of distress. DT scores correlated with HADS scores and ICQ subscales. No significant correlations were found between DT scores and clinical or demographical characteristics except for employment status. Conclusion. Prevalence of distress is high among patients with DTC even after long-term remission and cannot be predicted by clinical and demographical characteristics. DT and PL are useful screening instruments for distress in DTC patients and could easily be incorporated into daily practice.

Differentiated non-medullary thyroid carcinoma (DTC) is associated with a favorable prognosis with long-term survival rates that reach 80–95% [1]. Since most patients with DTC become in remission after treatment, the number of disease-free survivors is high. Quality of life (QoL) is an important issue in the care for long-term survivors of DTC [2]. Patients with DTC report impaired QoL in relation with initial surgery and hypothyroidism preceding treatment with radioactive iodine (RAI) [3,4]. Moreover, several studies have shown that even patients being in remission of DTC for many years report impaired QoL compared to healthy controls [4,5].

One factor that has negative effects on QoL is the level of distress. There is increasing evidence that cancer patients, including long-term survivors, are at increased risk of distress compared to healthy population controls [6–13], with distress defined as “an unpleasant experience of emotional, psychological, social or spiritual nature, that interferes with the ability to cope with cancer treatment, which extends along a continuum from common normal feelings of vulnerability, sadness and fear, to disabling problems such as true depression, anxiety, panic and feeling isolated or in a spiritual crisis” [10]. Distress has also been shown to negatively influence recovery time and treatment adherence in cancer patients [14]. Distress is often not noticed by medical professionals unless it is explicitly assessed [15,16]. For these reasons it is nowadays often recommended that all oncologic patients should be systematically screened for distress [9,17]. In order to meet the need for screening tools that can be used efficiently in an outpatient clinic setting, several short-tools have been developed for detection of distress. However, the majority of these have not been robustly validated in cancer settings [18].

The distress thermometer (DT) in combination with a problem list (PL) is a tool validated and currently recommended for screening for distress experienced by oncologic patients [19–21]. In contrast to other well-known QoL questionnaires, such as the hospital anxiety and depression scale (HADS), the self-reported problems on the PL are a starting point for further discussion in the clinical interview. In addition, the DT can be completed in just a few minutes, which makes it easy to implement in everyday outpatient practice. In this respect, a recent meta-analysis found the DT to be as accurate as the HADS as a short-tool to screen for distress in cancer settings but superior to it with respect to efficiency [22]. Although widely used in other types of cancer, the use of the DT has never been evaluated in patients with DTC.

Previous data suggest that patients with DTC perceive their illness as severe on a subjective, emotional basis unrelated to disease severity [23]. Therefore, the aim of this study was to evaluate the level of distress and problems experienced in a cohort of DTC subjects presenting for follow-up at the outpatient clinic of our tertiary referral academic hospital, using the DT and PL as short screening tools. In addition we investigated whether reported distress correlates with clinical, demographical and psychological variables measured by means of the HADS, illness cognition questionnaire (ICQ) and an ad hoc questionnaire in order to gain a broader view of the emotional and psychological factors influencing distress in DTC survivors.

Methods

Patients

All patients (>18 years old) who were treated for DTC and are under follow-up at the Department of Endocrinology of the Radboud University Nijmegen Medical Centre were invited to participate in the present study. The research has been approved by the Ethics Committee of the Radboud University Nijmegen Medical Centre. The patients were requested to fill in a questionnaire package sent by mail. Non-responders received a reminder letter four weeks later. Clinical variables were collected from medical charts and included information on diagnosis, treatment and follow-up. Primary treatment consisted of total thyroidectomy in all but two cases of papillary (PTC) microcarcinoma and one case of minimally invasive follicular thyroid carcinoma (FTC), and modified radical neck dissections in patients with confirmed nodal metastases. This was followed in the majority of patients by ablation with RAI (I¹³¹) of residual thyroid tissue four to six weeks after surgery. Initial cure was defined as undetectable thyroglobulin (Tg) in absence of anti-Tg antibodies and no evidence of loco-regional disease or distant metastases on whole body iodine scans (WBS) and/or neck ultrasonographic examinations at six months after RAI ablation. Tumor recurrence was defined as evidence of loco-regional disease or distant metastases more than six months after successful primary therapy. Current disease status was defined as cured (in remission) in case of undetectable Tg in the absence of anti-Tg antibodies and no evidence of loco-regional disease or distant metastases at last follow-up visit. Persistent disease was defined as either detectable Tg or evidence of loco-regional disease or distant metastases.

Questionnaires

Distress thermometer and problem list (DT and PL)

The DT is a modified visual analogue scale ranging from 0 (no distress) to 10 (extreme distress) resembling a thermometer [13]. Most of the validation studies in cancer patients have validated the DT against the HADS and established proper cut-off scores with receiver operating characteristics (ROC) curve analysis [6,11–13]. In the current study a ROC analysis was performed in order to establish the cut-off score for the DT using a HADS cut-off score ≥15 as the gold standard for detecting cases of severe emotional distress. This cut-off is similar to that used in a validation study performed in a cohort of patients with several other malignancies and having the same cultural background as our cohort [7,24].

We used the previously validated [7] Dutch version of the DT and problem list (PL) which is currently recommended for use in the clinical practice in the Netherlands. The Dutch version is adapted from the original national comprehensive cancer network (NCCN) version of the DT and PL based on evaluations by the Comprehensive Cancer Center the Netherlands (CCCN) focus groups. In comparison to the original 35 items of the original NCCN version, 15 items were added on the PL in the Dutch version (added items are depicted in italics in Table II). The items sadness, worry and loss of interest in usual activities were removed from the NCCN problem list, yielding a total of 47 items. The items are grouped in categories: physical problems (25), emotional problems (10), practical problems (7), social problems (3) and spiritual issues and religion (2). Respondents were instructed to indicate whether the items listed had been a problem in the past week by selecting from a fixed yes/no response [25]. In addition respondents were asked to indicate whether they would like to be referred to a professional for additional support [11].

Hospital anxiety and depression scale (HADS)

The HADS is a measure of anxiety and depression for patients with physical illness. It contains 14 items. Each item is scored between 0 and 3. Half of the items assess anxiety and the other half depression. For the anxiety and depression subscales scores 0–7 indicate normal values, 8–10 mild disorder, 11–14 moderate disorder and 15–21 severe disorder [26]. The two subscales can be combined into a single scale, and scores ≥15 on this scale were used to indicate severe emotional distress [7,24]. The HADS is widely used in oncology and has good reliability and validity [27]. The Dutch version of the HADS was used, which has been validated in several subgroups of Dutch patients [28].

Illness cognition questionnaire (ICQ)

The ICQ was used to measure helplessness, acceptance and perceived disease benefits. This is an 18-item questionnaire that contains three six-item scales, each with a scoring range of 6–24. Each item is answered on a four point scale to the extent to which one agrees with the item [29]. The ICQ was used because illness cognitions are related to psychological distress as theorized in Leventhal's Self-Regulatory Model (SRM) [30] and confirmed by several studies [31–33].

Demographic variables

Demographic information on age, gender, marital status, level of education, religion, age at diagnosis and employment status was collected by means of an ad-hoc questionnaire.

Statistical analysis

Spearman correlations were performed between the demographic characteristics, questionnaire scores and clinical characteristics. Significance was defined as p<0.05. ROC analysis was performed to examine the ability of the DT to detect distressed patients, and a HADS cut-off score ≥15 was set as the gold standard for detecting cases of clinically significant emotional distress [7,24]. Positive (PPV) and negative (NPV) predictive values were calculated for every DT score. All statistical analyses were also performed separately for the subgroup of patients currently being in remission of the disease according to the last follow-up data. All statistical analyses were performed using SPSS 16.0.

Results

Patient characteristics

Two hundred and five patients were eligible for this study. Of these, 159 patients responded (response rate 77.6%). Fourteen patients refused to participate because they had other interests or were unable to answer the questions because of language problems. A total of 145 patients [73.8% female, mean age (SD; range) 51.7 (13.8; 19–83) years] were enrolled in the study. Mean age at diagnosis was 40.1 (13.1; 9–79) years. Median follow-up time (range) since the last treatment was 7.2 years (3 months – 41 years). Regarding current disease status, most patients (118, 81.4%) were in remission and under follow-up, 18 (12.4%) patients had only detectable stable low Tg level without macroscopic evidence of disease and nine (6.2%) patients had persistent disease detectable on imaging. Serum TSH levels at last follow-up appointment had a mean (SD) value of 0.42 IU/L (1.12). Clinical characteristics are depicted in Table I. Participants and non-participants did not differ with respect to patient characteristics (p-values >0.2 for all items).

Table I. Patient characteristics and relation to distress thermometer (DT) score.

			Correlation to DT score
		%	Spearman's-rho	p-value
Gender (n): male/female	38/107		0.145	0.09
Age: mean (SD) (years)	51.7 (13.8)		−0.054	0.531
Age at diagnosis: mean (SD) (years)	40.1 (13.1)		−0.059	0.497
Duration since diagnosis: median (range) (years)	8.8 (0–41)		0.081	0.349
<1 year: n (%)	8 (5.5)
1–5 year: n (%)	29 (20.0)
5–10 year: n (%)	44 (30.3)
>10 year: n (%)	64 (44.1)
Follow-up since last treatment: median (range) (years)	7.2 (0–41)		0.036	0.676
Histopathological subtype: (n)
Papillary	106	73.1	−0.123	0.151
Folliculary	36	24.8	0.127	0.138
Mixed	3	2.1	0	0.994
TNM classification* stage: (n)			−0.094	0.292
I	68	46.9
II	22	15.2
III	11	7.6
IV	33	22.7
Unknown	11	7.6
Surgical treatment: (n)
Hemithyroidectomy	3	2.8	−0.128	0.14
Total thyroidectomy	140	96.6	0.052	0.55
Lymph node dissection	50	34.5	−0.133	0.125
Unknown	2	1.4
RAI treatment:
Number of treatments
0	12	8.3	0.067	0.437
1	76	52.4
2	37	25.5
>2	19	13.1
Unknown	1	0.7
Cumulative dose of I^131:
0 mCi	12	8.3	0.055	0.525
<100 mCi	6	4.1
100–200 mCi	70	48.3
>200 mCi	55	37.9
Unknown	2	1.4
Complications of surgery: (n)
Hypoparathyroidism (permanent)	43	29.7	0.134	0.118
Recurrent laryngeal nerve paralysis	8	5.5	0.023	0.786
No complications	95	65.5	−0.141	0.101
Disease status: (n)
Remission	110	80.3	0.054	0.533
Active disease	27	19.7
Mean TSH at last follow-up: (IU/L) (SD)	0.4 (1.1)		0.061	0.486

*UICC TNM Classification of malignant tumors, 6th edition [43].

Demographic characteristics

Two of the participants (1.4%) had less than eight years of education, 42 (29.2%) 8–12 years, 38 (26.4%) 12–16 years, 44 (30.6%) 16–21 years and 18 (12.5%) had a university degree. One participant omitted to fill in the highest education level. One hundred and twenty-five (86.2%) participants were married or engaged in a relationship and 20 (13.8%) had no relationship, were divorced or a widow/ widower. Eighty-two (56.6%) participants had full- or part-time employment, 34 (23.6%) were unemployed, 26 (18.1%) were retired and three (2.1%) were students.

Distress thermometer scores and problems

One hundred and thirty-seven patients rated their distress level on the DT and showed a mean score (SD) of 3.33 (2.70) with a 0–10 range. Of these, 47 participants (34.3%) rated their distress score ≥5, indicating clinically relevant distress as depicted later in the ROC analysis. Sixty participants (43.8%) rated their distress score >3 which is the NCCN distress management guideline cut-off [34]. Eight patients failed to score their distress on the DT because they forgot. Mean DT (SD) scores in the 110 patients being in remission and the 27 patients having persistent disease were 3.40 (2.72) and 3.04 (2.62) respectively and the percentages of patients reporting DT scores ≥5 in these groups were 36.4% and 25.9%.

There were no significant correlations between DT scores and the clinical characteristics regarding age at diagnosis, severity of the disease, disease status, treatments received or complications (Table I). In addition there was no correlation between the duration of (disease-free) follow-up or the TSH levels and DT scores. DT scores and level of education did not correlate significantly (r = −0.146, p= 0.089) in the total group. When analyzing the subgroup of patients being in remission of the disease, we found a slight but significant correlation between the DT score and the level of education (r = −0.212, p = 0.027). DT scores did not correlate significantly to having a partner (r = −0.072, p=0.402). There was a slight but significant correlation between the DT scores and the employment status (r = −0.251, p = 0.003), the participants having a full- or part-time employment reporting significantly less distress.

One hundred and forty-five respondents completed the PL (Table II). The most frequently reported problems were fatigue (62.1%), lack of physical fitness (46.9%), sleep problems (41.4%), muscle strength (40.0%) and nervousness (40.0%), weight changes (31.7%) and paresthesia (31.7%). The physical problems category was the most frequently reported category with 86.1% of the patients reporting at least one item. Emotional problems were reported by 76.0%, practical problems by 61.5%, social problems by 48.6% and problems regarding spiritual issues or religion by 39.4% of the patients. In total, 124 patients (89.9%) reported at least one item of concern on the PL. Of the patients reporting DT ≥5, 47 (100%) reported at least one item of concern on the PL. Of the patients with a DT score <5, 69 (76.7%) reported at least one item of concern on the PL.

Table II. Problem list (PL) items and correlation to distress thermometer (DT) score (*n represents the number of patients reporting a specific problem in the total group and in the active disease group).

				Correlation of total group PL score to total group DT score
Problem list	Mean score (SD)	n* (Total/active disease)	% (Total/active disease)	Spearman's-rho	p-value
Overall number of problems	9.5 (8.3)			0.827	<0.001
Practical problems	0.9 (1.3)			0.715	<0.001
Child care		13/2	9.0/7.4
Housing		8/1	5.5/3.7
House holding		41/6	28.3/22.2
Transportation		9/1	6.2/3.7
Work/school		35/9	24.1/33.3
Financial		18/4	12.4/14.8
Insurance		8/1	5.5/3.7
Social problems	0.4 (0.8)			0.468	<0.001
Dealing with partner		27/3	18.6/11.1
Dealing with children		20/4	13.8/14.8
Dealing with friends/relative		18/3	12.4/11.1
Emotional problems	2.5 (2.8)			0.652	<0.001
Grip on emotions		47/4	32.4/14.8
Memory		47/11	32.4/40.7
Confidence		35/4	24.1/14.8
Fears		31/4	21.4/14.8
Sadness		40/7	27.6/25.9
Nervousness		58/9	40.0/33.3
Loneliness		20/2	13.8/7.4
Concentration		51/8	35.2/29.6
Feeling of guilt		21/2	14.5/7.4
Loss of control		24/6	16.6/22.2
Spiritual-religious problems	0.2 (0.5)			0.233	0.006
Meaning of life		16/1	11.0/3.7
Faith in god/religion		13/5	9.0/18.5
Physical problems	5.3 (4.5)			0.782	<0.001
Appearance		22/5	15.2/18.5
Changes in urination		12/3	8.3/11.1
Constipation		23/2	15.9/7.4
Diarrhea		12/3	8.3/11.1
Eating		23/5	15.9/18.5
Feeling swollen		35/8	24.1/29.6
Fever		1/0	0.7/0.0
Mouth sores		18/8	12.4/29.6
Nausea		16/3	11.0/11.1
Nose dry/congested		38/4	26.2/14.8
Pain		35/5	24.1/18.5
Sexuality		25/4	17.2/11.1
Skin dry/itchy		42/5	29.0/18.5
Sleep		60/9	41.4/33.3
Breathing		17/2	11.7/7.4
Vertigo		28/4	19.3/14.8
Speech		17/3	11.7/11.1
Taste		14/5	9.7/18.5
Weight changes		46/11	31.7/40.7
Paresthesia in hands/feet		46/14	31.7/51.9
Washing/dressing		3/1	2.1/3.7
Getting around		26/5	17.9/18.5
Fatigue		90/15	62.1/55.6
Physical fitness		68/8	46.9/29.6
Muscle strength		58/5	40.0/18.5

There was a strong significant correlation between DT scores and the total number of reported problems on the PL (r = 0.827; p < 0.001). Significant correlations were also found between DT scores and the number of reported problems in each specific category. Table II shows the frequency distribution for endorsement of the PL and their correlation with the DT score.

HADS and ICQ

All participants completed the HADS questionnaire. The mean (SD) total score on the HADS was 9.25 (7.79), mean (SD) scores for anxiety and depression subscales were 5.40 (4.37) and 3.85 (3.93), respectively. Mean score on the HADS, number of patients reaching the subscale score ≥8 (threshold for mild, moderate and severe disorder) and their correlations with DT scores are depicted in Table III. Thirty-two (22.2%) of the patients had scores ≥15 (threshold for severe disorder).

Table III. Hospital anxiety and depression scale (HADS) and illness cognition questionnaire (ICQ) scores and correlation to distress thermometer (DT) scores (*n represents the number of patients with a score above the mentioned threshold).

		Correlation to DT score
		Spearman's-rho	p-value
HADS
Total	Mean score=9.2	0.611	<0.001
	SD=7.7
≥15 [n* (%)]	32 (22.1%)
Anxiety	Mean score=5.4	0.547	<0.001
	SD=4.3
≥8 [n* (%)]	36 (24.8%)
Depression	Mean score=3.8	0.602	<0.001
	SD=3.9
≥8 [n* (%)]	28 (19.3%)
ICQ
Helplessness	Mean score=9.0	0.662	<0.001
	SD=3.6
Acceptance	Mean score=19.0	−0.544	<0.001
	SD=4.2
Disease benefits	Mean score=16.1	−0.092	0.285
	SD=4.5

The mean (SD) scores on helplessness, acceptance and disease benefits domains of the ICQ were 9.03 (3.66), 19.08 (4.28) and 16.13 (4.52), respectively. There was a significant correlation between DT score and helplessness and acceptance domains (Table III). One participant failed to fully complete the ICQ.

Cut-off scores and predictive values

The ROC curve predicting elevated distress according to the HADS showed an area under the curve of 0.82 (standard error 0.043; 95% confidence interval, 0.739–0.906; p<0.001) (Figure 1). Table IV lists the sensitivity and specificity values for all DT scores, including frequencies. The cut-off score of 5 correctly identified 77% of HADS cases (sensitivity) and 77% of HADS non-cases (specificity).

Figure 1. Receiver operating characteristics (ROC) curve of distress thermometer (DT) scores versus hospital anxiety and depression scale (HADS) cut-off scores. AUC 0.82, SD 0.043, 95% CI 0.739–0.906, p < 0.001.

Table IV. Frequencies, sensitivity, specificity, positive predictive values (PPV), and negative predictive values (NPV) of distress thermometer (DT) scores (*n represents the number of patients with the respective score in the total group and in the group of patients with active disease).

DT score	n* (cum.%) Total group	n* (cum.%) Active disease	Sensitivity	Specificity	PPV	NPV
0	26 (19.1)	6 (22.2)	1.00	0.00	0.22	0.00
1	18 (32.4)	5 (40.7)	0.97	0.24	0.26	0.96
2	17 (44.9)	1 (44.4)	0.97	0.41	0.32	0.98
3	15 (55.9)	4 (59.3)	0.90	0.55	0.36	0.95
4	13 (65.4)	4 (74.1)	0.80	0.66	0.40	0.92
5	12 (74.3)	0 (74.1)	0.77	0.77	0.49	0.92
6	12 (83.1)	3 (85.2)	0.60	0.84	0.51	0.88
7	13 (92.6)	3 (96.3)	0.50	0.92	0.65	0.87
8	7 (97.8)	1 (100.0)	0.20	0.96	0.60	0.81
9	2 (99.3)	0 (100.0)	0.07	0.99	0.67	0.77
10	1(100.0)	0 (100.0)	0.04	1.00	1.00	0.79

Wish for referral

Nineteen patients (13.1%) expressed their wish to discuss with another professional for additional support. Thirty-seven patients (25.5%) answered this question with “maybe”. Of the patients reporting a DT score ≥5, 19.1% wished to be referred for additional support and 34.0% were considering it (and answered “maybe”). There was a significant correlation between the DT score and the wish for referral for additional support (r = 0.193, p = 0.025). “Maybe” willing to be referred for additional support significantly correlated with DT score as well (r = 0.230, p = 0.007). Four respondents omitted to report whether they wish to be referred.

Discussion

Cancer patients experience high levels of distress [6,10,20]. Relying on the subjective estimates of distress levels by medical professionals is often inaccurate and misleading [15,16]. The DT combined with a PL is a tool developed initially by NCCN [34] for screening for distress in patients with breast tumors. It has been validated for use in several types of cancer and in populations with different ethnic backgrounds, including a series of Dutch patients [6,7,11,13,35]. This easy-to-use instrument helps identify treatable problems and specific issues that may cause distress, allowing physicians to tackle these issues more targeted. In addition, patients are explicitly requested whether they wish to be referred for additional specialized help. This potentially improves communication between patients and healthcare providers as well as patient care and satisfaction and may help to use limited consultation time more effectively [8]. This is the first study evaluating the level of distress using the DT and PL in DTC patients.

The main finding of the present study is that the patients with DTC report experiencing significant levels of distress even years after being cured, with 34.3% of the patients reporting distress scores ≥5 on the DT, which may count for significant distress and possibly requires intervention. Surprisingly, this percentage is comparable to the 28.6–50% reported in series [19,36–39] with similar response rates, including patients with other types of cancer such as lung-, breast cancer and leukemia, having a more aggressive course and being associated with more physical burden than patients with DTC, the majority of whom were in long-term remission. However, these studies differ with respect to type of cancer, patient characteristics, cultural background, type of treatment, duration of follow-up and DT cut-off level [4–6] used to define distress, and therefore cannot be compared appropriately. In DTC patients, although on the long-term QoL approaches that of the healthy population, previous studies indicate that most long-term survivors report specific persistent problems such as the negative effects of thyroid hormone withdrawal for radioiodine follow-up procedures, the fear and uncertainty related to a cancer diagnosis, feelings of diagnosis being dismissed as not seriously or “having a benign cancer”, confrontation with daily use of thyroid hormone medication and the fluctuating thyroid hormone levels [4]. The nature and intensity of these problems can change during the oncologic trajectory and become sources of distress. As the duration of follow-up increases for instance, the thyroid hormone withdrawal is not required anymore and the thyroid hormone levels become stable and remain within the normal range in the majority of patients. On the other hand, other problems such as concerns about fertility can become more relevant after longer follow-up. Therefore, identifying patients who may experience distress and may benefit from additional help and addressing the specific problems should be important objectives in the ongoing care of these long-term DTC survivors. When considering the range of DT scores and the nature of reported problems we conclude that the DT is a useful screening tool for this particular population.

In consistence with findings in other types of cancer [19], no significant correlations were found between level of distress and either clinical or demographic characteristics, with the exception of employment status. The role of employment status in patients with chronic conditions and in oncologic patients is not clear and data in the literature are lacking. It has been suggested that unemployment influences distress levels in patients with cardiovascular disease [40]. Although not completely comparable with DTC survivors, this might support the need for reintegration strategies for patients in the work process after they have been cured. This is particularly important for patients with DTC who are often very young at time of diagnosis. Some specific problems reported by patients in this study slightly correlated with clinical characteristics, such as paresthesia in patients with hypoparathyroidism or mouth sores and impaired speech in patients receiving a higher cumulative dose of RAI (data not shown). These problems are very specific for thyroid carcinoma patients and should be recognized and addressed by the physicians. The majority of reported problems however did not correlate with clinical characteristics. Altogether these results demonstrate that distress is a profound individual reaction which cannot be predicted by objective clinical data and emphasize the need for routine specific screening for distress.

Consistent with previous findings, patients cited physical problems over emotional problems as sources of distress [41]. Nevertheless, the correlation of DT scores with those of HADS and ICQ helplessness domain suggest that physical problems may lead to an increased psychological burden for these patients. In addition, the negative correlation with scores of the ICQ acceptance domain suggests that the increasing burden renders the patients less likely to accept these problems.

The high DT score alone does not automatically imply that the patient should be referred for additional support, despite the clear correlation between DT score and the wish to be referred. Many problems that can be important sources of distress can be resolved by discussing them with nurses or physicians, others may require consultation of other professionals such as dieticians, physical therapists or social workers. The PL that accompanies the DT is a very useful starting point when discussing with patients about their specific problems. Previous data suggest that acknowledging and discussing problems alone may already yield clinical benefits and improve patient satisfaction [19]. In addition, the wish for intervention for individual patients must be taken into account, and therefore some patients can explicitly refuse further referral or just “maybe” consider it. In this series, although majority of patients did not express a wish to be referred for additional support, 13.1% of them addressed wishes for referral and 25.5% “maybe” considered it. These percentages are lower compared with other forms of cancer such as lung cancer where approximately 22% expressed wishes for referral [20]. Nevertheless, considering that this series consisted mainly of long-term disease-free patients, these numbers are relevant and the needs of these patients should be addressed in a multidisciplinary fashion.

Distress scores correlated with HADS anxiety, depression and total scores. This is in accordance with previous data in other types of cancer [42]. Cut-off score of 5 resulted in optimal sensitivity and specificity relative to the HADS. The sensitivity of the DT in the current study was comparable to the results of a recent meta-analysis (77% vs. 78.3% in the meta-analysis), whereas specificity was higher (77% vs. 66.5% in the meta-analysis) [22]. In addition, NPV in the current study (92%) was higher compared to the meta-analysis (84%) [22], confirming that the DT is excellent for ruling out clinically elevated distress in this patient category, but given its low PPV of 49% it is less useful as diagnostic tool. Therefore a DT score ≥5 is only a starting point for further diagnostic evaluation by a physician or specialized nurse, if necessary followed by further intervention.

This study has a number of limitations. The cross sectional retrospective nature of this study does not allow evaluation of the course of distress in time. Some periods during treatment or follow-up might be associated with higher levels of distress, such as the period short after establishing the diagnosis and the period of invasive treatment (surgery, RAI). Also patients with persistent disease might experience more distress than long-term disease-free survivors. This series includes only few patients with persistent disease, as expected given the good prognosis of the disease, and lacks the statistical power required to analyze these correlations. In addition this study was not designed to investigate the clinical benefit of screening these patients.

Nevertheless, this study represents the first evaluation of the level of distress in patients with DTC. It includes a large number of patients with DTC with long-term remission, who represent the majority of patients seen on the oncologic endocrine follow-up clinic. In addition, it is the first study to evaluate the validity of the DT accompanied by the PL against the HADS for screening of distress in this patient category.

In conclusion, the prevalence of distress is high in patients with DTC even after long-term remission. Physical and emotional problems were the main sources of distress. Clinical and socio-demographic data were not correlated with distress in DTC patients, except for the employment status. The DT and PL are useful time efficient screening instruments for psychosocial distress in DTC patients and could easily be incorporated into daily practice. These findings highlight the importance of a routine psychological distress screening even in a later phase after treatment. Future research should focus on the benefits of prospective and systematic screening for distress in patients with DTC, as well as best intervention strategies.

Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.