- Email: [email protected]
Knowledge of cancer symptoms and anxiety affect patient delay in seeking diagnosis in patients with heterogeneous cancer locations
Curr Probl Cancer ] (2016) ]]]–]]]
Contents lists available at ScienceDirect
Curr Probl Cancer journal homepage: www.elsevier.com/locate/cpcancer
Knowledge of cancer symptoms and anxiety affect patient delay in seeking diagnosis in patients with heterogeneous cancer locations Gabriela Chojnacka-Szawłowska, PhD, Mikołaj Majkowicz, PhD, Krzysztof Basiński, MA, Agata Zdun-Ryżewska, PhD, Iwona Wasilewko, PhD, Piotr Pankiewicz, PhD a bs t r a c t Purpose: This research was aimed at identifying factors that predict patient delay in treatment initiation in patients with suspected cancer disease. We sought to determine the differences between delaying and nondelaying patients with reference to their knowledge of cancer symptoms, sociodemographic variables, and the levels of state anxiety and trait anxiety. Methods: The study involved 301 randomly selected patients with suspected cancer disease before their first oncology appointment at a regional oncology center in Poland. Data were collected by means of a semistructured interview conducted by a trained psychologist. To evaluate the knowledge of cancer symptoms, the symptoms mentioned by subjects were compared to the list of symptoms from cancer awareness measure. Anxiety levels were assessed using the State-Trait Anxiety Inventory. Results: In the course of logistic regression analysis a model was developed, in which knowledge of cancer symptoms and state anxiety allowed to predict patient delay. Knowledge of every additional cancer symptom decreased the chance of patient delay by 16.4% point [95% CI: 1.4–29.2]. An increase in state anxiety for every point of the scale decreased the chance of delay by 2.5% points [95% CI: 0.2–4.6]. Trait anxiety and the studied sociodemographic variables proved to be nonsignificant predictors of patient delay. Conclusions: Knowledge of cancer symptoms and the level of state anxiety allowed to predict patient delay in the initiation of treatment. Owing to the heterogeneity of the tumor locations within the sample, the obtained model can be used in large scale prevention programs designed for the whole population. & 2016 Elsevier Inc. All rights reserved.
This work was supported by the Medical University of Gdańsk statutory funds under Grant ST 275279. http://dx.doi.org/10.1016/j.currproblcancer.2016.10.001 0147-0272/& 2016 Elsevier Inc. All rights reserved.
G. Chojnacka-Szawłowska et al. / Curr Probl Cancer ] (2016) ]]]–]]]
2
Introduction Diagnosing cancer at an early stage, and thereby reducing mortality, is the goal of tertiary oncological prevention.1 At an onset of unusual symptoms, patient delay in treatment initiation poses a major problem to cancer treatment. The time span in such cases is usually measured in months or even years between noticing the unusual symptoms and seeking doctor’s advice. The process of delaying is psychologically complex and its explanation involves mental properties and social characteristics of an individual, as well as his or her awareness of ongoing psychophysiological processes.2 Patient-related delaying dramatically increases the number of patients initiating treatment at an advanced stage of a disease.3 Postponing diagnostic tests means avoiding a psychologically difficult situation, when a person does not want to learn about the state of their health. This kind of avoidance can be generalized to all health-related behaviors.4 The aim of our research was to identify the psychological and sociodemographic variables responsible for patient delay, regardless of a tumor location. The empirical part the research is a continuation of a broader project on different aspects of patient delay determinants.20
Table 1 Descriptive statistics of the studied variables in groups with delays and no delays in seeking cancer treatment. Significant differences between groups (P o 0.05) in bold. Total (N ¼ 283)
Delay (N ¼ 180)
No delay (N ¼ 103)
P
Age M (795% CI) SD
45.8 (44.1-47.5) 14.2
47.1 (44.9-49.2) 14.5
43.7 (41.1-46.3) 13.4
0.06a
Sex Women, % (N) Men, % (N)
62.9 (178) 37.1 (105)
57.2 (103) 42.8 (77)
72.8 (75) 27.2 (28)
0.009b
Educational level, % (N) Low Medium High
31.1 (88) 59.0 (167) 9.9 (28)
33.3 (60) 57.8 (104) 8.9 (16)
27.2 (28) 61.2 (63) 11.7 (12)
0.49b
Marital status, % (N) Single Married, informal relationship Divorced, widow(er), other
11.7 (33) 80.2 (227) 8.1 (23)
10.6 (19) 81.1 (146) 8.3 (15)
13.6 (14) 78.6 (81) 7.8 (8)
0.59b
Place of residence, % (N) Urban Rural
73.1 (207) 26.9 (76)
72.2 (130) 27.8 (50)
74.8 (77) 25.2 (26)
0.21b
Knowledge of symptoms M (795% CI) SD
1.6 (1.4-1.7) 1.5
1.3 (1.1-1.5) 1.3
1.9 (1.6-2.3) 1.8
0.01a
State anxiety, M (SD) M (795% CI) SD
42.7 (41.1-44.3) 13.3
40.7 (38.8-42.7) 13.1
46.0 (43.5-48.5) 13.0
0.002a
Trait anxiety, M (SD) M (795% CI) SD
41.2 (40.1-42.3) 9.3
40.9 (39.5-42.4) 9.8
41.6 (40.0-43.2) 8.4
0.16a
a b
Mann-Whitney U. Pearson chi-squared.
G. Chojnacka-Szawłowska et al. / Curr Probl Cancer ] (2016) ]]]–]]]
3
Materials and methods Participants The study involved 301 patients with suspected cancer, who arrived for a preliminary examination at their first oncological consultation. Every third person claiming the appointment was asked to participate in the research. When a request met with a refusal, the subsequent third person was chosen. The survey was conducted before the oncological appointment so as to eliminate any confounders (eg, new information about health and a prospect of stressful diagnostic tests). The suspicion of cancer was verified post hoc in the following numbers (N) of patients: breast cancer, N ¼ 140; bowel cancer, N ¼ 29; bronchial carcinoma, N ¼ 28; reproductive organs cancers, N ¼ 16; lymphomas, N ¼ 15; cancers of head and neck, N ¼ 15; gastric carcinoma, N ¼ 14; prostatic carcinoma, N ¼ 8; sarcoma, N ¼ 6; laryngeal cancer, N ¼ 6; and urinary system cancer, N ¼ 4. Other forms of cancer were suspected in single individuals. After tests (radiology, histopathology, blood tests, and others), organ-specified location and clinical severity of the disease were determined. In 10 patients the diagnostic process was either not complete or unclear. In 35 patients cancer disease was not confirmed, and in 48 patients the tumor was benign. The research was conducted in the Regional Oncology Centre of Gdansk, Poland. The results of 18 persons have not been included in the analysis owing to missing data in State-Trait Anxiety Inventory (STAI). Eventually, the sample included N ¼ 283 individuals with demographic characteristics presented in Table 1. The mean age was 45.7 years (standard deviation [SD] ¼ 14.1). Women accounted for 62.9% of the sample. Among the participants, 31.1% had primary education, 59% secondary or higher education, and 9.9% university education. People in stable relationships accounted for 80.2%, 11.7% participants were single, and 8.1% were divorced or in other relationships. In towns or cities or villages lived 73.1% and 26.9% participants, respectively. Methods Data on patient delay in seeking oncological advice, data on cancer symptom awareness, and demographic data were collected using an interview conducted by a trained psychologist. Interviewees were asked when they had first noticed the unusual symptoms and what happened subsequently. This part of the interview was unstructured so as to help participants speak freely, feeling no fear of any kind of judgment on the part of the interviewer. Knowledge of cancer symptoms was evaluated in the structured part of the interview. Respondents were asked to name symptoms of cancer that were known to them. Subsequently, the interviewer compared the list of mentioned symptoms with the list of 10 cancer symptoms taken from the cancer awareness measure (CAM).21,22 The CAM list comprises the following symptoms: persistent cough or hoarseness; persistent change in bowel habits; unexplained persistent pain; persistent change in bladder habits; unexplained lump; a change in the appearance of a mole; a sore that does not heal; unexplained bleeding; unexplained weight loss; or persistent difficulty swallowing. An interviewee received 1 point for mentioning any symptom from the CAM list. The general indicator of knowledge for each evaluated subject was obtained by adding points. The study applied the State-Trait Anxiety Inventory (STAI), which was designed to measure individual anxiety level differences in response to various threats and stressors. The tool comprises 2 subscales, which enable to determine the anxiety level as a relatively constant trait of an individual and as a state in which the tested person presently is.23 Statistical analyses All statistical analyses were performed with STATISTICA v.12 (StatSoft Inc). The statistical significance of between-group differences (delay vs no delay) was estimated by nonparametric
4
G. Chojnacka-Szawłowska et al. / Curr Probl Cancer ] (2016) ]]]–]]]
Mann-Whitney U test for continuous variables or χ2 for categorical variables. Binomial logistic regression was performed to evaluate the effect of anxiety level and knowledge of cancer symptoms on patient delay. To evaluate the significance of sociodemographic variables (ie, sex, age, education, place of residence, and marital status) 2 models were constructed. The first model included only the sociodemographic variables, whereas the second model was complemented with indicators of anxiety and knowledge of cancer symptoms. The results are presented as adjusted odds ratios with 95% CI.
Theory Several studies, with their origins dating back as far as the first half of the 20th century, presented complex determinants of patient delay and led to an understanding of various psychological factors, most notably anxiety and fatalism,5 unrealistic optimism,6 and negative affect.7 The importance of sociodemographic variables such as young age,8 poor education,9 and poor knowledge of cancer symptoms10 was also stressed. In the delay process, these variables coexist in various relationships, depending on the group of patients under examination. With certain cancers (eg, ovarian), the clinical limitations of symptoms evaluation also play an important role in delays, when the symptoms are long present, discrete, difficult to identify, or change with the disease progression.11 Numerous inconsistencies among the research outcomes inhibit synthesis and affect applying the knowledge into prevention practice. Such state of research is partially a consequence of conducting multifaceted studies on patients with homogeneous cancer location and applying various research methodologies. Thus, paradoxically, the extensive knowledge does not translate into cancer prevention and patient delay in treatment initiation continues to present a problem. Relatively little is known about relations among knowledge of cancer symptoms, anxiety, beliefs about treatment efficacy, and initiating diagnostic tests.12 Research outcomes revealed that poor understanding of symptoms was associated with insufficient participation in screening tests13 and interpreting symptoms as signs of oncological disease increased anxiety about the diagnosis.14 Prolonged patient delay, sometimes measured in years, was associated with anxiety and fatalistic beliefs about cancer incurability,15 which could lead to defensive denial or to ignoring symptoms16 by attributing them to trivial reasons and interpreting them as harmless.17 Symptoms were often treated as natural, not health-relevant consequences of hormonal changes or injuries.18,19
Results Table 1 presents demographic characteristics of the sample along with the anxiety and knowledge of cancer symptoms measures across groups with and without delay in seeking treatment. Sex appeared significantly associated with delay in initiating treatment. In the delaying group, women accounted for 57.2%, compared with 72.8 % in nondelaying; χ2 ¼ 6.549, P o 0.01. The association measured by Cramer0 s V coefficient appeared weak, with V ¼ 0.148, P o 0.01. Featured groups differed also in the knowledge of cancer symptoms. In the nondelaying group, the knowledge of symptoms was significantly higher (M ¼ 1.9, SD ¼ 1.8) than in the delaying (M ¼ 1.3, SD ¼ 1.3). The most frequently mentioned symptoms included were: unexplained lump (65.2%), a change in the appearance of a mole (20.5%), and persistent change in bowel habits (12.6%). As for the level of state anxiety, the nondelaying patients scored significantly higher (M ¼ 46.0; SD ¼ 13.0) than the delaying (M ¼ 40.7; SD ¼ 13.1). For other tested variables, statistically significant differences were not detected. It is of note that anxiety levels were higher in women in comparison with men. In women, mean state anxiety was M ¼ 46.2, SD ¼ 13.2 and mean trait anxiety was M ¼ 43.2, SD ¼ 9.2. In
G. Chojnacka-Szawłowska et al. / Curr Probl Cancer ] (2016) ]]]–]]]
5
Table 2 Logistic regression of delays in seeking cancer treatment (delay ¼ 1, no delay ¼ 0). Coefficients are adjusted odds ratios. Significant estimates (P o 0.05) in bold. Predictor
Adjusted OR
95% CI
P
Age Sex (woman)
1.007 0.678
0.986-1.029 0.374-1.229
0.50 0.20
Educational level (reference: low) Medium High
0.882 0.725
0.474-1.641 0.282-1.863
0.69 0.51
Marital status (reference: single) Married, informal relationship Divorced, widow(er), other Place of residence (urban) Knowledge of symptoms
1.150 1.216 0.857 0.836
0.497-2.662 0.343-4.315 0.474-1.615 0.708-0.986
0.74 0.76 0.67 0.034
Anxiety State Trait
0.975 1.011
0.954-0.998 0.978-1.044
0.03 0.53
OR, odds ratios.
men, these measures were M ¼ 36.7, SD ¼ 11.2 and M ¼ 37.8, and SD ¼ 8.4, respectively. Both comparisons yielded statistically significant differences (Mann-Whitney U, P o 0.001 for state anxiety, and P o 0.001 for trait anxiety). To determine the capacity of the studied variables to predict patient delay in seeking treatment, 2 logistic regression models were constructed. The model that assumed an effect of demographic variables (ie, age, sex, education, marital status, and place of residence) proved statistically nonsignificant (Omnibus Test, P ¼ 0.194). In the other model, knowledge of cancer symptoms was analyzed alongside demographic variables and the values of state anxiety and trait anxiety (Table 2). That model proved statistically significant (P o 0.05). Knowledge of cancer symptoms and state anxiety were significant predictors of patient delay. For knowledge of cancer symptoms, the odds ratio was 0.836 (95% CI: 0.708-0.986; P o 0.05). Awareness of each additional symptom of the disease decreased the chance of patient delay by 16.4% points, regardless of the values of other variables. The odds ratio for state anxiety was 0.975 (95% CI: 0.954-0.998; P o 0.05). For each scale point, the increase in state anxiety decreased the chance of patient delay by 2.5% points, regardless of the values of other variables. Trait anxiety was a nonsignificant predictor in the model. In comparison with the null model, which assumed that all respondents would postpone their visit, the model assuming an effect of anxiety and knowledge of cancer symptoms allowed for better predictions of the delay (65.4% vs 63.6% accurate predictions). What should be noted is the large disparity between the percentages of correct predictions of delaying and nondelaying patients (89.4% and 23.3%, respectively).
Discussion The results of our research demonstrate a possibility of predicting patient delay in oncological examinations, where patient delay is related to knowledge of cancer symptoms and state anxiety. In this study, the more cancer symptoms the examined person was aware of and the higher was the state anxiety, the lesser was the chance of delay. In such relationship, complexly interrelated cognitive and emotional processes mediate in giving meaning to symptoms and influence the rise of anxiety, thus relating to the time of patient delay.24
6
G. Chojnacka-Szawłowska et al. / Curr Probl Cancer ] (2016) ]]]–]]]
Similarly to our results, data from other studies confirm the link between better understanding of cancer symptoms and shorter patient delay.16,19,25,26 Other studies have also shown that lack of knowledge of cancer symptoms was associated with incorrect interpretation of symptoms and longer delay in claiming medical appointments.27,28 According to the results of our research, anxiety can be seen as a signal of an impending danger, appearing in response to symptoms seen as threatening regarding cognition and behavior. From the cognitive perspective, anxiety facilitates an early detection of a danger signal, and from the behavioral perspective, it enables mobilizing resources against the coming events, thus avoiding danger.29 Within the research sample, the percentage of delaying women was lower than nondelaying. However, sex proved an insignificant predictor in the regression model, owing most probably to the fact that state anxiety was the mediator of the dependence. The state anxiety values were significantly higher in women than in men. Variables and relationships described by the proposed model appear to be simple and the outcomes are useful in practice. Moreover, merely 2 variables, knowledge of cancer symptoms and state anxiety, allow to predict patient delay, regardless of cancer location. However, difficult it is to predict human behavior in life-threatening circumstances, the results of our research show that accurate predicting is possible in a situation as difficult as undergoing oncological diagnostic tests. The research results are important in the field of cancer prevention and useful in the course of diagnosing tumors of heterogeneous locations. They highlight the importance of knowledge about nonspecific cancer symptoms and the mobilizing role of state anxiety in initiating treatment. Oncological prevention that increases awareness of different cancer symptoms, simultaneous with a situationally triggered natural anxiety, can contribute toward shorter patient delay. For this reason, educational measures should be doubly aimed at the necessity of increasing awareness of specific and nonspecific symptoms of heterogeneously located cancers and at emphasizing the importance of early treatment initiation to make the treatment less burdensome and improve the chance of recovery.
Strengths and limitations The obtained model allows predictions of patient delay, raising the prospect of eliminating the important and difficult problem of oncological prevention. Its insufficiency for accurate predictions of nondelaying, owing most likely to the interference of variables not controlled in the research, does not limit the scope of effective preventive applications.
Conclusions The results of this research show that better knowledge of cancer symptoms and higher state anxiety decrease the possibility of patient delay in initiating treatment. The presented model can be applied in tertiary oncological prevention aimed at the general population. References 1. Andersen BL. Introduction to the featured section: psychological and behavioral studies in cancer prevention and control. Heal Psychol 1996;15(5):411–412. 2. Andersen BL, Cacioppo JT, Roberts DC. Delay in seeking a cancer diagnosis: delay stages and psychophysiological comparison processes. Br J Soc Psychol 1996;34(March):33–52. 3. Montgomery M. Uncertainty during breast diagnostic evaluation: state of the science. Oncol Nurs Forum 2010;37(1): 77–83. 4. Li WWY, Lam WWT, Wong JHF, et al. Waiting to see the doctor: understanding appraisal and utilization components of consultation delay for new breast symptoms in Chinese women. Psychooncology 2012;21(12):1316–1323. 5. Peek ME, Sayad JV, Markwardt R. Fear, fatalism and breast cancer screening in low-income African-American women: the role of clinicians and the health care system. J Gen Intern Med 2008;23(11):1847–1853.
G. Chojnacka-Szawłowska et al. / Curr Probl Cancer ] (2016) ]]]–]]]
7
6. Gold RS. The link between judgments of comparative risk and own risk: further evidence. Psychol Health Med 2007;12(2):238–247. 7. Lyubomirsky S, Kasri F, Chang O, Chung I. Ruminative response styles and delay of seeking diagnosis for breast cancer symptoms. J Soc Clin Psychol 2006;25(3):276–304. 8. Meyer BJF, Talbot AP, Ranalli C. Why older adults make more immediate treatment decisions about cancer than younger adults. Psychol Aging 2007;22(3):505–524. 9. Ristvedt SL, Trinkaus KM. Sex differences in responding to rectal cancer symptoms. Psychol Health 2008;23(8): 935–944. 10. Tod AM, Craven J, Allmark P. Diagnostic delay in lung cancer: a qualitative study. J Adv Nurs 2008;61(3):336–343. 11. Koldjeski D, Kirkpatrick MK, Swanson M, Everett L, Brown S. An ovarian cancer diagnosis-seeking process: unraveling the diagnostic delay problem. Oncol Nurs Forum 2005;32(5):1036–1042. 12. de Nooijer J, Lechner L, de Vries H. Help-seeking behaviour for cancer symptoms: perceptions of patients and general practitioners. Psychooncology 2001;10(6):469–478. 13. Othman AK, Kiviniemi MT, Wu YWB, Lally RM. Influence of demographic factors, knowledge, and beliefs on Jordanian women0 s intention to undergo mammography screening. J Nurs Scholarsh 2012;44(1):19–26. 14. Forbes LJL, Warburton F, Richards MA, Ramirez AJ. Risk factors for delay in symptomatic presentation: a survey of cancer patients. Br J Cancer 2014;111:581–588. 15. Phelan M, Dobbs J, David AS. ‘I thought it would go away’: patient denial in breast cancer. J R Soc Med 1992;85(4): 206–207. 16. McCutchan GM, Wood F, Edwards A, Richards R, Brain KE. Influences of cancer symptom knowledge, beliefs and barriers on cancer symptom presentation in relation to socioeconomic deprivation: a systematic review. BMC Cancer 2015;15:1000. 17. Smith LK, Pope C, Botha JL. Patients’ help-seeking experiences and delay in cancer presentation: a qualitative synthesis. Lancet 2005;366(9488):825–831. 18. Khakbazan Z, Taghipour A, Latifnejad Roudsari R, Mohammadi E. Help seeking behavior of women with selfdiscovered breast cancer symptoms: a meta-ethnographic synthesis of patient delay. PLoS One 2014;9:e110262. 19. Quaife SL, Forbes LJL, Ramirez AJ, et al. Recognition of cancer warning signs and anticipated delay in help-seeking in a population sample of adults in the UK. Br J Cancer 2014;110(1):12–18. 20. Chojnacka-Szawłowska G, Kościelak R, Karasiewicz K, Majkowicz M, Kozaka J. Delays in seeking cancer diagnosis in relation to beliefs about the curability of cancer in patients with different disease locations. Psychol Health 2013;28 (2):154–170. 21. Stubbings S, Robb K, Waller J, et al. Development of a measurement tool to assess public awareness of cancer. Br J Cancer 2009;101(Suppl):13–17. 22. Whitaker KL, Winstanley K, Macleod U, Scot SE, Wardle J. Low cancer suspicion following experience of a cancer ‘warning sign0 . Eur J Cancer 2015;51(16):2473–2479. 23. Spielberger CD. State-Trait Anxiety Inventory. Palo Alto: Consulting Pschychologists Press; 1983. 24. Facione NC, Facione PA. The cognitive structuring of patient delay in breast cancer. Soc Sci Med 2006;63(12): 3137–3149. 25. Robb K, Stubbings S, Ramirez A, et al. Public awareness of cancer in Britain: a population-based survey of adults. Br J Cancer 2009;101(suppl 2):S18–S23. 26. Simon AE, Waller J, Robb K, Wardle J. Patient delay in presentation of possible cancer symptoms: the contribution of knowledge and attitudes in a population sample from the United kingdom. Cancer Epidemiol Biomarkers Prev 2010;19(9):2272–2277. 27. Rauscher GH, Ferrans CE, Kaiser K, Campbell RT, Calhoun EE, Warnecke RB. Misconceptions about breast lumps and delayed medical presentation in urban breast cancer patients. Cancer Epidemiol Biomarkers Prev 2010;19(3):640–647. 28. Ristvedt SL, Trinkaus KM. Psychological factors related to delay in consultation for cancer symptoms. Psychooncology 2005;14(5):339–350. 29. Castillo MD, Calvo MG. Anxiety gives priority to anticipation of threatening events. Eur Psychol 2000;5(3):234.
Contents lists available at ScienceDirect
Curr Probl Cancer journal homepage: www.elsevier.com/locate/cpcancer
Knowledge of cancer symptoms and anxiety affect patient delay in seeking diagnosis in patients with heterogeneous cancer locations Gabriela Chojnacka-Szawłowska, PhD, Mikołaj Majkowicz, PhD, Krzysztof Basiński, MA, Agata Zdun-Ryżewska, PhD, Iwona Wasilewko, PhD, Piotr Pankiewicz, PhD a bs t r a c t Purpose: This research was aimed at identifying factors that predict patient delay in treatment initiation in patients with suspected cancer disease. We sought to determine the differences between delaying and nondelaying patients with reference to their knowledge of cancer symptoms, sociodemographic variables, and the levels of state anxiety and trait anxiety. Methods: The study involved 301 randomly selected patients with suspected cancer disease before their first oncology appointment at a regional oncology center in Poland. Data were collected by means of a semistructured interview conducted by a trained psychologist. To evaluate the knowledge of cancer symptoms, the symptoms mentioned by subjects were compared to the list of symptoms from cancer awareness measure. Anxiety levels were assessed using the State-Trait Anxiety Inventory. Results: In the course of logistic regression analysis a model was developed, in which knowledge of cancer symptoms and state anxiety allowed to predict patient delay. Knowledge of every additional cancer symptom decreased the chance of patient delay by 16.4% point [95% CI: 1.4–29.2]. An increase in state anxiety for every point of the scale decreased the chance of delay by 2.5% points [95% CI: 0.2–4.6]. Trait anxiety and the studied sociodemographic variables proved to be nonsignificant predictors of patient delay. Conclusions: Knowledge of cancer symptoms and the level of state anxiety allowed to predict patient delay in the initiation of treatment. Owing to the heterogeneity of the tumor locations within the sample, the obtained model can be used in large scale prevention programs designed for the whole population. & 2016 Elsevier Inc. All rights reserved.
This work was supported by the Medical University of Gdańsk statutory funds under Grant ST 275279. http://dx.doi.org/10.1016/j.currproblcancer.2016.10.001 0147-0272/& 2016 Elsevier Inc. All rights reserved.
G. Chojnacka-Szawłowska et al. / Curr Probl Cancer ] (2016) ]]]–]]]
2
Introduction Diagnosing cancer at an early stage, and thereby reducing mortality, is the goal of tertiary oncological prevention.1 At an onset of unusual symptoms, patient delay in treatment initiation poses a major problem to cancer treatment. The time span in such cases is usually measured in months or even years between noticing the unusual symptoms and seeking doctor’s advice. The process of delaying is psychologically complex and its explanation involves mental properties and social characteristics of an individual, as well as his or her awareness of ongoing psychophysiological processes.2 Patient-related delaying dramatically increases the number of patients initiating treatment at an advanced stage of a disease.3 Postponing diagnostic tests means avoiding a psychologically difficult situation, when a person does not want to learn about the state of their health. This kind of avoidance can be generalized to all health-related behaviors.4 The aim of our research was to identify the psychological and sociodemographic variables responsible for patient delay, regardless of a tumor location. The empirical part the research is a continuation of a broader project on different aspects of patient delay determinants.20
Table 1 Descriptive statistics of the studied variables in groups with delays and no delays in seeking cancer treatment. Significant differences between groups (P o 0.05) in bold. Total (N ¼ 283)
Delay (N ¼ 180)
No delay (N ¼ 103)
P
Age M (795% CI) SD
45.8 (44.1-47.5) 14.2
47.1 (44.9-49.2) 14.5
43.7 (41.1-46.3) 13.4
0.06a
Sex Women, % (N) Men, % (N)
62.9 (178) 37.1 (105)
57.2 (103) 42.8 (77)
72.8 (75) 27.2 (28)
0.009b
Educational level, % (N) Low Medium High
31.1 (88) 59.0 (167) 9.9 (28)
33.3 (60) 57.8 (104) 8.9 (16)
27.2 (28) 61.2 (63) 11.7 (12)
0.49b
Marital status, % (N) Single Married, informal relationship Divorced, widow(er), other
11.7 (33) 80.2 (227) 8.1 (23)
10.6 (19) 81.1 (146) 8.3 (15)
13.6 (14) 78.6 (81) 7.8 (8)
0.59b
Place of residence, % (N) Urban Rural
73.1 (207) 26.9 (76)
72.2 (130) 27.8 (50)
74.8 (77) 25.2 (26)
0.21b
Knowledge of symptoms M (795% CI) SD
1.6 (1.4-1.7) 1.5
1.3 (1.1-1.5) 1.3
1.9 (1.6-2.3) 1.8
0.01a
State anxiety, M (SD) M (795% CI) SD
42.7 (41.1-44.3) 13.3
40.7 (38.8-42.7) 13.1
46.0 (43.5-48.5) 13.0
0.002a
Trait anxiety, M (SD) M (795% CI) SD
41.2 (40.1-42.3) 9.3
40.9 (39.5-42.4) 9.8
41.6 (40.0-43.2) 8.4
0.16a
a b
Mann-Whitney U. Pearson chi-squared.
G. Chojnacka-Szawłowska et al. / Curr Probl Cancer ] (2016) ]]]–]]]
3
Materials and methods Participants The study involved 301 patients with suspected cancer, who arrived for a preliminary examination at their first oncological consultation. Every third person claiming the appointment was asked to participate in the research. When a request met with a refusal, the subsequent third person was chosen. The survey was conducted before the oncological appointment so as to eliminate any confounders (eg, new information about health and a prospect of stressful diagnostic tests). The suspicion of cancer was verified post hoc in the following numbers (N) of patients: breast cancer, N ¼ 140; bowel cancer, N ¼ 29; bronchial carcinoma, N ¼ 28; reproductive organs cancers, N ¼ 16; lymphomas, N ¼ 15; cancers of head and neck, N ¼ 15; gastric carcinoma, N ¼ 14; prostatic carcinoma, N ¼ 8; sarcoma, N ¼ 6; laryngeal cancer, N ¼ 6; and urinary system cancer, N ¼ 4. Other forms of cancer were suspected in single individuals. After tests (radiology, histopathology, blood tests, and others), organ-specified location and clinical severity of the disease were determined. In 10 patients the diagnostic process was either not complete or unclear. In 35 patients cancer disease was not confirmed, and in 48 patients the tumor was benign. The research was conducted in the Regional Oncology Centre of Gdansk, Poland. The results of 18 persons have not been included in the analysis owing to missing data in State-Trait Anxiety Inventory (STAI). Eventually, the sample included N ¼ 283 individuals with demographic characteristics presented in Table 1. The mean age was 45.7 years (standard deviation [SD] ¼ 14.1). Women accounted for 62.9% of the sample. Among the participants, 31.1% had primary education, 59% secondary or higher education, and 9.9% university education. People in stable relationships accounted for 80.2%, 11.7% participants were single, and 8.1% were divorced or in other relationships. In towns or cities or villages lived 73.1% and 26.9% participants, respectively. Methods Data on patient delay in seeking oncological advice, data on cancer symptom awareness, and demographic data were collected using an interview conducted by a trained psychologist. Interviewees were asked when they had first noticed the unusual symptoms and what happened subsequently. This part of the interview was unstructured so as to help participants speak freely, feeling no fear of any kind of judgment on the part of the interviewer. Knowledge of cancer symptoms was evaluated in the structured part of the interview. Respondents were asked to name symptoms of cancer that were known to them. Subsequently, the interviewer compared the list of mentioned symptoms with the list of 10 cancer symptoms taken from the cancer awareness measure (CAM).21,22 The CAM list comprises the following symptoms: persistent cough or hoarseness; persistent change in bowel habits; unexplained persistent pain; persistent change in bladder habits; unexplained lump; a change in the appearance of a mole; a sore that does not heal; unexplained bleeding; unexplained weight loss; or persistent difficulty swallowing. An interviewee received 1 point for mentioning any symptom from the CAM list. The general indicator of knowledge for each evaluated subject was obtained by adding points. The study applied the State-Trait Anxiety Inventory (STAI), which was designed to measure individual anxiety level differences in response to various threats and stressors. The tool comprises 2 subscales, which enable to determine the anxiety level as a relatively constant trait of an individual and as a state in which the tested person presently is.23 Statistical analyses All statistical analyses were performed with STATISTICA v.12 (StatSoft Inc). The statistical significance of between-group differences (delay vs no delay) was estimated by nonparametric
4
G. Chojnacka-Szawłowska et al. / Curr Probl Cancer ] (2016) ]]]–]]]
Mann-Whitney U test for continuous variables or χ2 for categorical variables. Binomial logistic regression was performed to evaluate the effect of anxiety level and knowledge of cancer symptoms on patient delay. To evaluate the significance of sociodemographic variables (ie, sex, age, education, place of residence, and marital status) 2 models were constructed. The first model included only the sociodemographic variables, whereas the second model was complemented with indicators of anxiety and knowledge of cancer symptoms. The results are presented as adjusted odds ratios with 95% CI.
Theory Several studies, with their origins dating back as far as the first half of the 20th century, presented complex determinants of patient delay and led to an understanding of various psychological factors, most notably anxiety and fatalism,5 unrealistic optimism,6 and negative affect.7 The importance of sociodemographic variables such as young age,8 poor education,9 and poor knowledge of cancer symptoms10 was also stressed. In the delay process, these variables coexist in various relationships, depending on the group of patients under examination. With certain cancers (eg, ovarian), the clinical limitations of symptoms evaluation also play an important role in delays, when the symptoms are long present, discrete, difficult to identify, or change with the disease progression.11 Numerous inconsistencies among the research outcomes inhibit synthesis and affect applying the knowledge into prevention practice. Such state of research is partially a consequence of conducting multifaceted studies on patients with homogeneous cancer location and applying various research methodologies. Thus, paradoxically, the extensive knowledge does not translate into cancer prevention and patient delay in treatment initiation continues to present a problem. Relatively little is known about relations among knowledge of cancer symptoms, anxiety, beliefs about treatment efficacy, and initiating diagnostic tests.12 Research outcomes revealed that poor understanding of symptoms was associated with insufficient participation in screening tests13 and interpreting symptoms as signs of oncological disease increased anxiety about the diagnosis.14 Prolonged patient delay, sometimes measured in years, was associated with anxiety and fatalistic beliefs about cancer incurability,15 which could lead to defensive denial or to ignoring symptoms16 by attributing them to trivial reasons and interpreting them as harmless.17 Symptoms were often treated as natural, not health-relevant consequences of hormonal changes or injuries.18,19
Results Table 1 presents demographic characteristics of the sample along with the anxiety and knowledge of cancer symptoms measures across groups with and without delay in seeking treatment. Sex appeared significantly associated with delay in initiating treatment. In the delaying group, women accounted for 57.2%, compared with 72.8 % in nondelaying; χ2 ¼ 6.549, P o 0.01. The association measured by Cramer0 s V coefficient appeared weak, with V ¼ 0.148, P o 0.01. Featured groups differed also in the knowledge of cancer symptoms. In the nondelaying group, the knowledge of symptoms was significantly higher (M ¼ 1.9, SD ¼ 1.8) than in the delaying (M ¼ 1.3, SD ¼ 1.3). The most frequently mentioned symptoms included were: unexplained lump (65.2%), a change in the appearance of a mole (20.5%), and persistent change in bowel habits (12.6%). As for the level of state anxiety, the nondelaying patients scored significantly higher (M ¼ 46.0; SD ¼ 13.0) than the delaying (M ¼ 40.7; SD ¼ 13.1). For other tested variables, statistically significant differences were not detected. It is of note that anxiety levels were higher in women in comparison with men. In women, mean state anxiety was M ¼ 46.2, SD ¼ 13.2 and mean trait anxiety was M ¼ 43.2, SD ¼ 9.2. In
G. Chojnacka-Szawłowska et al. / Curr Probl Cancer ] (2016) ]]]–]]]
5
Table 2 Logistic regression of delays in seeking cancer treatment (delay ¼ 1, no delay ¼ 0). Coefficients are adjusted odds ratios. Significant estimates (P o 0.05) in bold. Predictor
Adjusted OR
95% CI
P
Age Sex (woman)
1.007 0.678
0.986-1.029 0.374-1.229
0.50 0.20
Educational level (reference: low) Medium High
0.882 0.725
0.474-1.641 0.282-1.863
0.69 0.51
Marital status (reference: single) Married, informal relationship Divorced, widow(er), other Place of residence (urban) Knowledge of symptoms
1.150 1.216 0.857 0.836
0.497-2.662 0.343-4.315 0.474-1.615 0.708-0.986
0.74 0.76 0.67 0.034
Anxiety State Trait
0.975 1.011
0.954-0.998 0.978-1.044
0.03 0.53
OR, odds ratios.
men, these measures were M ¼ 36.7, SD ¼ 11.2 and M ¼ 37.8, and SD ¼ 8.4, respectively. Both comparisons yielded statistically significant differences (Mann-Whitney U, P o 0.001 for state anxiety, and P o 0.001 for trait anxiety). To determine the capacity of the studied variables to predict patient delay in seeking treatment, 2 logistic regression models were constructed. The model that assumed an effect of demographic variables (ie, age, sex, education, marital status, and place of residence) proved statistically nonsignificant (Omnibus Test, P ¼ 0.194). In the other model, knowledge of cancer symptoms was analyzed alongside demographic variables and the values of state anxiety and trait anxiety (Table 2). That model proved statistically significant (P o 0.05). Knowledge of cancer symptoms and state anxiety were significant predictors of patient delay. For knowledge of cancer symptoms, the odds ratio was 0.836 (95% CI: 0.708-0.986; P o 0.05). Awareness of each additional symptom of the disease decreased the chance of patient delay by 16.4% points, regardless of the values of other variables. The odds ratio for state anxiety was 0.975 (95% CI: 0.954-0.998; P o 0.05). For each scale point, the increase in state anxiety decreased the chance of patient delay by 2.5% points, regardless of the values of other variables. Trait anxiety was a nonsignificant predictor in the model. In comparison with the null model, which assumed that all respondents would postpone their visit, the model assuming an effect of anxiety and knowledge of cancer symptoms allowed for better predictions of the delay (65.4% vs 63.6% accurate predictions). What should be noted is the large disparity between the percentages of correct predictions of delaying and nondelaying patients (89.4% and 23.3%, respectively).
Discussion The results of our research demonstrate a possibility of predicting patient delay in oncological examinations, where patient delay is related to knowledge of cancer symptoms and state anxiety. In this study, the more cancer symptoms the examined person was aware of and the higher was the state anxiety, the lesser was the chance of delay. In such relationship, complexly interrelated cognitive and emotional processes mediate in giving meaning to symptoms and influence the rise of anxiety, thus relating to the time of patient delay.24
6
G. Chojnacka-Szawłowska et al. / Curr Probl Cancer ] (2016) ]]]–]]]
Similarly to our results, data from other studies confirm the link between better understanding of cancer symptoms and shorter patient delay.16,19,25,26 Other studies have also shown that lack of knowledge of cancer symptoms was associated with incorrect interpretation of symptoms and longer delay in claiming medical appointments.27,28 According to the results of our research, anxiety can be seen as a signal of an impending danger, appearing in response to symptoms seen as threatening regarding cognition and behavior. From the cognitive perspective, anxiety facilitates an early detection of a danger signal, and from the behavioral perspective, it enables mobilizing resources against the coming events, thus avoiding danger.29 Within the research sample, the percentage of delaying women was lower than nondelaying. However, sex proved an insignificant predictor in the regression model, owing most probably to the fact that state anxiety was the mediator of the dependence. The state anxiety values were significantly higher in women than in men. Variables and relationships described by the proposed model appear to be simple and the outcomes are useful in practice. Moreover, merely 2 variables, knowledge of cancer symptoms and state anxiety, allow to predict patient delay, regardless of cancer location. However, difficult it is to predict human behavior in life-threatening circumstances, the results of our research show that accurate predicting is possible in a situation as difficult as undergoing oncological diagnostic tests. The research results are important in the field of cancer prevention and useful in the course of diagnosing tumors of heterogeneous locations. They highlight the importance of knowledge about nonspecific cancer symptoms and the mobilizing role of state anxiety in initiating treatment. Oncological prevention that increases awareness of different cancer symptoms, simultaneous with a situationally triggered natural anxiety, can contribute toward shorter patient delay. For this reason, educational measures should be doubly aimed at the necessity of increasing awareness of specific and nonspecific symptoms of heterogeneously located cancers and at emphasizing the importance of early treatment initiation to make the treatment less burdensome and improve the chance of recovery.
Strengths and limitations The obtained model allows predictions of patient delay, raising the prospect of eliminating the important and difficult problem of oncological prevention. Its insufficiency for accurate predictions of nondelaying, owing most likely to the interference of variables not controlled in the research, does not limit the scope of effective preventive applications.
Conclusions The results of this research show that better knowledge of cancer symptoms and higher state anxiety decrease the possibility of patient delay in initiating treatment. The presented model can be applied in tertiary oncological prevention aimed at the general population. References 1. Andersen BL. Introduction to the featured section: psychological and behavioral studies in cancer prevention and control. Heal Psychol 1996;15(5):411–412. 2. Andersen BL, Cacioppo JT, Roberts DC. Delay in seeking a cancer diagnosis: delay stages and psychophysiological comparison processes. Br J Soc Psychol 1996;34(March):33–52. 3. Montgomery M. Uncertainty during breast diagnostic evaluation: state of the science. Oncol Nurs Forum 2010;37(1): 77–83. 4. Li WWY, Lam WWT, Wong JHF, et al. Waiting to see the doctor: understanding appraisal and utilization components of consultation delay for new breast symptoms in Chinese women. Psychooncology 2012;21(12):1316–1323. 5. Peek ME, Sayad JV, Markwardt R. Fear, fatalism and breast cancer screening in low-income African-American women: the role of clinicians and the health care system. J Gen Intern Med 2008;23(11):1847–1853.
G. Chojnacka-Szawłowska et al. / Curr Probl Cancer ] (2016) ]]]–]]]
7
6. Gold RS. The link between judgments of comparative risk and own risk: further evidence. Psychol Health Med 2007;12(2):238–247. 7. Lyubomirsky S, Kasri F, Chang O, Chung I. Ruminative response styles and delay of seeking diagnosis for breast cancer symptoms. J Soc Clin Psychol 2006;25(3):276–304. 8. Meyer BJF, Talbot AP, Ranalli C. Why older adults make more immediate treatment decisions about cancer than younger adults. Psychol Aging 2007;22(3):505–524. 9. Ristvedt SL, Trinkaus KM. Sex differences in responding to rectal cancer symptoms. Psychol Health 2008;23(8): 935–944. 10. Tod AM, Craven J, Allmark P. Diagnostic delay in lung cancer: a qualitative study. J Adv Nurs 2008;61(3):336–343. 11. Koldjeski D, Kirkpatrick MK, Swanson M, Everett L, Brown S. An ovarian cancer diagnosis-seeking process: unraveling the diagnostic delay problem. Oncol Nurs Forum 2005;32(5):1036–1042. 12. de Nooijer J, Lechner L, de Vries H. Help-seeking behaviour for cancer symptoms: perceptions of patients and general practitioners. Psychooncology 2001;10(6):469–478. 13. Othman AK, Kiviniemi MT, Wu YWB, Lally RM. Influence of demographic factors, knowledge, and beliefs on Jordanian women0 s intention to undergo mammography screening. J Nurs Scholarsh 2012;44(1):19–26. 14. Forbes LJL, Warburton F, Richards MA, Ramirez AJ. Risk factors for delay in symptomatic presentation: a survey of cancer patients. Br J Cancer 2014;111:581–588. 15. Phelan M, Dobbs J, David AS. ‘I thought it would go away’: patient denial in breast cancer. J R Soc Med 1992;85(4): 206–207. 16. McCutchan GM, Wood F, Edwards A, Richards R, Brain KE. Influences of cancer symptom knowledge, beliefs and barriers on cancer symptom presentation in relation to socioeconomic deprivation: a systematic review. BMC Cancer 2015;15:1000. 17. Smith LK, Pope C, Botha JL. Patients’ help-seeking experiences and delay in cancer presentation: a qualitative synthesis. Lancet 2005;366(9488):825–831. 18. Khakbazan Z, Taghipour A, Latifnejad Roudsari R, Mohammadi E. Help seeking behavior of women with selfdiscovered breast cancer symptoms: a meta-ethnographic synthesis of patient delay. PLoS One 2014;9:e110262. 19. Quaife SL, Forbes LJL, Ramirez AJ, et al. Recognition of cancer warning signs and anticipated delay in help-seeking in a population sample of adults in the UK. Br J Cancer 2014;110(1):12–18. 20. Chojnacka-Szawłowska G, Kościelak R, Karasiewicz K, Majkowicz M, Kozaka J. Delays in seeking cancer diagnosis in relation to beliefs about the curability of cancer in patients with different disease locations. Psychol Health 2013;28 (2):154–170. 21. Stubbings S, Robb K, Waller J, et al. Development of a measurement tool to assess public awareness of cancer. Br J Cancer 2009;101(Suppl):13–17. 22. Whitaker KL, Winstanley K, Macleod U, Scot SE, Wardle J. Low cancer suspicion following experience of a cancer ‘warning sign0 . Eur J Cancer 2015;51(16):2473–2479. 23. Spielberger CD. State-Trait Anxiety Inventory. Palo Alto: Consulting Pschychologists Press; 1983. 24. Facione NC, Facione PA. The cognitive structuring of patient delay in breast cancer. Soc Sci Med 2006;63(12): 3137–3149. 25. Robb K, Stubbings S, Ramirez A, et al. Public awareness of cancer in Britain: a population-based survey of adults. Br J Cancer 2009;101(suppl 2):S18–S23. 26. Simon AE, Waller J, Robb K, Wardle J. Patient delay in presentation of possible cancer symptoms: the contribution of knowledge and attitudes in a population sample from the United kingdom. Cancer Epidemiol Biomarkers Prev 2010;19(9):2272–2277. 27. Rauscher GH, Ferrans CE, Kaiser K, Campbell RT, Calhoun EE, Warnecke RB. Misconceptions about breast lumps and delayed medical presentation in urban breast cancer patients. Cancer Epidemiol Biomarkers Prev 2010;19(3):640–647. 28. Ristvedt SL, Trinkaus KM. Psychological factors related to delay in consultation for cancer symptoms. Psychooncology 2005;14(5):339–350. 29. Castillo MD, Calvo MG. Anxiety gives priority to anticipation of threatening events. Eur Psychol 2000;5(3):234.