JDR JDR Most Cited Articles
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (16)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Pitiphat, W.
Right arrow Articles by Zavras, A.I.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Pitiphat, W.
Right arrow Articles by Zavras, A.I.
J Dent Res 81(3): 192-197, 2002
© 2002 International and American Associations for Dental Research

RESEARCH REPORT
Clinical

Factors Associated with Delay in the Diagnosis of Oral Cancer

W. Pitiphat1,2,3, S.R. Diehl4, G. Laskaris5, V. Cartsos1, C.W. Douglass1,2, and A.I. Zavras1,*

1 Department of Oral Health Policy & Epidemiology, Harvard School of Dental Medicine, 188 Longwood Ave., Boston, MA 02115, USA;
2 Department of Epidemiology, Harvard School of Public Health, Boston, MA 02115, USA;
3 Department of Community Dentistry, Faculty of Dentistry, Khon Kaen University, Khon Kaen 40002, Thailand;
4 Craniofacial Epidemiology and Genetics Branch, National Institute of Dental and Craniofacial Research, Bethesda, MD 20892-6401, USA; and
5 Oral Medicine Department, A. Sygros Hospital, University of Athens Medical School, 37 Ipsiladou Street, Athens 10676, Greece;

*corresponding author, zavras{at}hms.harvard.edu


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS & METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Early detection and treatment improve the prognosis for oral cancer. Delays from the onset of symptoms to clinical diagnosis are common. Our aim is to identify factors associated with this delay. Between 1995 and 1998, we interviewed 105 consecutive patients with histologically confirmed oral cancer in Greece. If 21 or more days elapsed from the time the patient noticed major symptoms to a definitive diagnosis, we called it a delay (52% of cases). We used logistic and linear regression to estimate odds ratios of delayed diagnosis and to identify correlates of length of delay, respectively. Former smokers had a 4.3 times greater risk of delayed diagnosis compared with current smokers (95% confidence interval: 1.1-17.1). The length of delay was greater among single patients, non-smokers, or those with stage IV tumors. Clinicians should be advised that delay in the diagnosis of oral cancer occurs frequently, even in individuals who do not smoke heavily.

KEY WORDS: diagnosis • delay • oral cavity • mouth neoplasms


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS & METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Oral cancer has one of the lowest five-year survival rates among the major types of cancer, including breast, skin, testis, prostate, uterus, and urinary bladder cancers (Pisani et al., 1999), with survival rates often of 50% or less (Ries et al., 2000). Early diagnosis is crucial to an improved survival rate. If lesions are detected when they are small, localized, and treated expeditiously, survival rates of 70 to 90% can be achieved (Silverman, 1988). The five-year survival rate for persons with localized lesions is four times greater than that for those with distant metastases (Ries et al., 2000).

Although oral cancer occurs in a part of the body that is readily accessible for early detection, most lesions are not diagnosed until they have reached advanced stages. Based on the Surveillance, Epidemiology, and End Results (SEER) data, at the time of diagnosis of oral cancer, 36% of persons had localized disease, 44% had regional disease, and 9% had distant disease (Ries et al., 2000).

Previous studies in several populations have shown that there is often a substantial delay in the diagnosis of oral cancer (Dimitroulis et al., 1992; Wildt et al., 1995; Hollows et al., 2000). For earlier diagnosis to be promoted, it is important that the factors predisposing to diagnostic delays be identified. We therefore examined the extent and determinants of delay in the diagnosis of oral cancer in the Greek population.


   MATERIALS & METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS & METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Study Participants and Data Collection
We evaluated factors associated with diagnostic delay among the oral cancer cases in a hospital-based case-control study in Greece (Zavras et al., 2001). The study was conducted in 3 major teaching hospitals in Athens, between November, 1995, and January, 1998. Patients were adults aged 26 to 91 years, with no prior history of oral cancerC, who were diagnosed with histopathologically confirmed squamous cell oral or pharyngeal cancer (ICD9 141, 143-145, 148-149). Informed consent was obtained from all patients who participated in the study. The protocol was reviewed and approved by Institutional Review Boards of both the Harvard School of Public Health and the National Institute of Dental and Craniofacial Research (NIDCR/NIH).

Patients were interviewed in the hospital by a trained interviewer using a structured, pre-tested, questionnaire. Risk factor data included demographic and socio-economic characteristics, information on tobacco use, alcohol drinking, family history of cancer, intra-oral status, and weight change. Tumor size and TNM stage at time of diagnosis were also assessed. We recorded the time interval from the self-reported date when oral cancer signs and/or symptoms were first noted by the cases to the date of definitive diagnosis. Some of the most common symptoms reported by cases included prolonged hemorrhage, pain, existence of a tumor/nodule, and difficulty in swallowing. A subject was defined as being delayed if 21 or more days had elapsed between first notice of signs or symptoms by the subject and the definitive oral cancer diagnosis. This interval of 3 wks allows for a seven- to 10-day follow-up of a symptom, a second visit and a biopsy, as well as the time required for a histopathologist to report the results back to the dentist/physician. For those subjects who were coded as being delayed, the length of delay was defined as the number of days beyond the 20th day after initial symptoms were self-recognized to the date of the definitive oral cancer diagnosis. For example, if a subject reported that the interval between first notice of symptoms and definite diagnosis was 90 days, then his/her length of delay was 90 – 20 = 70 days.

Statistical Analysis
We first analyzed the data by comparing the distributions of study variables between delay and non-delay groups. We evaluated the statistical significance of differences between the two groups (time to diagnosis and other continuous variables) using the Wilcoxon rank-sum test. Odds ratios (OR) and 95% confidence intervals (CI) were estimated by logistic regression. We utilized multiple linear regression to identify factors associated with the length of delay among those subjects who experienced delay of >= 21 days. The following variables were considered: gender, age, education (no formal education, 1-6 yrs, 7-12 yrs, > 12 yrs of schooling), marital status (single, married, widowed, separated), employment status (having being unemployed or not), history of liver cirrhosis (yes, no), history of sexually transmitted disease (yes, no), family history of cancer (yes, no), alcohol drinking (never or less than 1 glass/wk, 1-28 glasses/wk, 29-42 glasses/wk, > 42 glasses/wk), number of missing teeth, weight change, tumor size, and stage of tumor (advanced = TNM stage IV; earlier = TNM stage I, II, III). Cigarette smoking was analyzed both as discrete categories (non-smoker, former smoker, current smoker) and as continuous exposure (pack yrs). All statistical tests were two-tailed.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS & METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
During the study period, 105 respondents provided complete information on the variables of interest. The ratio of males to females was 1.6:1. Self-reported time to diagnosis varied from 0 to 780 days, with a mean of 80.6 days (SD 157.9) and a median of 30 days. Fifty-five patients exhibited a delay of 21 days or more (52.4%). Of these patients with a delay in diagnosis, 21 (38.2%) were diagnosed more than 14 wks after their first symptoms, and 11 (20.0%) more than 30 wks after first symptoms (Fig.Go).



View larger version (22K):
[in this window]
[in a new window]
  		Figure. Time to diagnosis for 105 oral and pharyngeal cancer cases in Greece.

 
Most patients were married and employed, and had no history of liver cirrhosis or sexually transmitted disease. More than 60% of the cases were not diagnosed until they reached TNM stage II or beyond. Delay and non-delay groups were similar in most characteristics (Table 1Go). There was no significant difference in the risk of a diagnostic delay between former smokers and non-smokers, and between current smokers and non-smokers. The risk was significantly higher among former smokers than among current smokers (OR, 4.3; 95%CI, 1.1-17.1). No other characteristics were associated with delay status at significance level of 0.05.


View this table:
[in this window]
[in a new window]
  		Table 1. Study Characteristics of the Patients According to Status of Delay in the Diagnosis of Oral Cancers
 
The univariate linear regression analyses performed for the identification of factors related to the length of delay (among those cases with 21 or more days' delay in diagnosis) are presented in Table 2Go. Length of delay was significantly longer in unmarried than in ever-married patients, and in non-smokers than in ever-smokers. Delay time was marginally associated with gender and advanced stage of tumor. Weight loss was significantly and positively related to the length of delay (p < 0.01), but only in the univariate analyses.


View this table:
[in this window]
[in a new window]
  		Table 2. Simple Linear Regression Analyses of the Association between the Total Delay Time to Diagnosis and Patient's Characteristics (n = 55)
 
In the multivariate linear regression analyses (Table 3Go), the length of delay in diagnosis was significantly associated with marital status, tobacco use, and advanced stage of tumor. The diagnosis was delayed approximately 239 days for patients with TNM stage IV as compared with those with earlier stages of disease. Past and current smokers were diagnosed on an average of 133 and 131 days earlier than non-smokers, respectively. The longest diagnostic delay was associated with being unmarried, since these subjects experienced delays estimated to be 170, 383, and 248 days longer than patients who were married, widowed, and separated, respectively.


View this table:
[in this window]
[in a new window]
  		Table 3. Multiple Linear Regression Analysis to Identify Predictors of Delay Days in Diagnosis
 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS & METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
Fifty-two percent of Greek patients with oral cancer had a delay in diagnosis of more than 3 wks. The time from initial symptoms to definitive diagnosis ranged between 0 and 780 days, with a median of 30 days. A study in The Netherlands reported similar findings (median time of 46 days, range 14 days to 2 yrs) (Jovanovic et al., 1992). A median time to diagnosis of 3 mos was observed in Canada (Elwood and Gallagher, 1985) and Italy (Mashberg et al., 1989), compared with 4 mos in Finland (Söderholm, 1990), Denmark (Wildt et al., 1995), and Israel (Gorsky and Dayan, 1995). Although our results are more favorable compared with those of other populations, the proportion of patients with delay was still considerably high.

A previous study showed no association between marital status and oral cancer delay in diagnosis (Elwood and Gallagher, 1985). In our study, unmarried patients experienced longer delay than ever-married subjects. Being unmarried has been reported as associated with less favorable life-styles and unhealthy behaviors. Unmarried persons may be less likely to have dental care coverage (Manski, 1995) and less likely to utilize dental services (Osterberg et al., 1998). Patients with newly diagnosed and recurrent cancer who are married appear to experience higher levels of hope, as assessed by the Herth Hope Scale (Ballard et al., 1997). Two large longitudinal studies conducted in the US (Johnson et al., 2000) and the UK (Cheung, 2000) showed that unmarried persons had an elevated risk of death compared with married persons.

Our finding of an absence of significant association between age and diagnostic delay is consistent with findings from other studies in oral cancer patients (Elwood and Gallagher, 1985; Guggenheimer et al., 1989; Jovanovic et al., 1992) and in head and neck cancer patients in general (Amir et al., 1999). One study showed a significant but weak correlation between female gender (r = 0.26) and older age (r = 0.19) with professional delay (interval between first visit until definitive diagnosis), but not with patient delay (interval between onset of symptoms to first visit to a physician) (Wildt et al., 1995). A recent study of tongue cancer found that fatal delays too often occurred when the initial professional evaluation did not lead to a follow-up referral for further examination (Kantola et al., 2001). In Thailand, prolonged patient delay was associated with the use of traditional herbal medicine (Kerdpon and Sriplung, 2001), highlighting how local social and cultural factors need to be taken into consideration for different populations.

We found no association between gender and education and delay in diagnosis. Similar results have been reported by various workers (Guggenheimer et al., 1989; Jovanovic et al., 1992; Amir et al., 1999); one study observed a longer time to diagnosis in females than in males, but these results were only marginally significant (p = 0.05) (Elwood and Gallagher, 1985). Surrogate measures for socio-economic status, such as education level and unemployment, were found not to affect the timing to diagnosis. Since it is known that socio-economic status (SES) affects the five-year survival rate, as is the case with US blacks as compared with whites, residual confounding cannot be ruled out. Our study focused on patients being diagnosed at public hospitals of the National Healthcare System. A future study is needed to assess those who are diagnosed or treated at private hospitals.

Our findings suggest a strong association between history of sexually transmitted disease and delay in diagnosis. History of a sexually transmitted disease may be associated with personal behaviors that were not assessed in the present work, such as low healthcare services utilization. For the exact nature of this interesting association to be assessed, more studies are needed. These studies need to control for residual confounding, in addition to having adequate statistical power.

Theoretically, one would expect that longer delay would be associated with more advanced stages of disease at diagnosis. We confirmed this theory only for TNM stage IV. In one study, patients with TI cancers had a shorter delay than patients with larger lesions, but patients with TII, TIII, and TIV cancers were similar in length of delay (Mashberg et al., 1989). Other studies indicated no association of tumor staging with overall delay (Guggenheimer et al., 1989; Söderholm, 1990; Jovanovic et al., 1992; Kowalski et al., 1994), patient delay (Kowalski et al., 1994; Wildt et al., 1995; Hollows et al., 2000), or professional delay (Wildt et al., 1995; Hollows et al., 2000). Apart from possible population differences, the discrepancies in these results may be explained in several ways. First, we performed a multiple linear regression using length of delay as the dependent variable, while most of the above-mentioned studies performed bivariate categorical analyses, comparing the distribution of tumor staging among categories of number of wks or mos of delay. Classifying length of delay into categories may be subject to misclassification, leading to null results. Second, we focused our analyses on the subgroup of patients who had delay in diagnosis, because our main research question was to identify factors that influence the length of delay. To determine the robustness of our method, we also performed the analyses on all our study patients (data not shown), and found that the results were similar, although the association between Stage IV and length of delay became only marginally significant (p = 0.06).

The absence of a significant association between delay in diagnosis and alcohol use is in agreement with the literature (Elwood and Gallagher, 1985; Guggenheimer et al., 1989; Hollows et al., 2000). Previous studies also found no association between delay and smoking habit (Elwood and Gallagher, 1985; Hollows et al., 2000). We found a significant relationship of delay with smoking history, but not with the quantity consumed, with former smokers being more likely to have delay in diagnosis as compared with current smokers. Among patients with delay, non-smokers had a longer delay than both former and current smokers. It is possible that being a non-smoker may generate a false feeling of security for both the patient and the physician. Another possible explanation is that smokers might seek professional attention more often than non-smokers (Rodriguez Artalejo et al., 2000), due to a variety of chronic medical, dental, and smoking-related conditions, thereby resulting in an earlier diagnosis of oral cancer.

Assessment of the time to diagnosis depended on patients' recollection of their first symptoms, and thus, this is a limitation of our study. It is possible that subjects did not correctly recall the onset of symptoms, especially when the delayed period was very prolonged. Although there may be measurement error in estimates of the length of delay, we doubt that there is substantial misclassification of our subjects as being delayed or not, since the median number of days of delay differ so dramatically between these groups (90 vs. 0 days). Another limitation is our inability to distinguish if the delay is due to the patient or the clinician, and thus, to infer the true causes of observed delays.

Our study confirms previous international reports regarding the magnitude of the problem and sets the stage for additional investigations of the reasons for diagnostic delays and the potential for the development of effective strategies for reducing this serious problem. Such strategies should consider the long delays observed in subjects who are not exposed to high-risk behaviors such as smoking and heavy drinking. The fact that delay was associated with the stage with the highest morbidity and mortality, Stage IV oral cancer, should alert clinicians and the public about the value of early detection. Rigorous educational programs targeting both the general public and health professionals (Yellowitz et al., 2000), supplemented by innovative diagnostic strategies, such as the use of tolonium chloride for high-risk groups (Kerawala et al., 2000) or the use of brush biopsy cytological detection of abnormalities (Sciubba, 1999), promise to lower the burden of oral cancer.


   ACKNOWLEDGMENTS
 
The authors acknowledge the support of Drs. J. Segas, P. Nomikos, D. Lefantzis, G. Dokianakis, and P. Thomas. Special thanks to A. Pantelidakis for serving as interviewer, to Dr. Mohammad Khoshnevisan, Dr. Kaumudi Joshipura, and Renukha Bahadursingh for review and editorial assistance in the preparation of the manuscript. This study was partially supported by NIDCR/NIH K23DE00420 grant and by NIDR contracts MD 626829 and 726655.

Received April 27, 2001; Last revision November 16, 2001; Accepted January 16, 2002


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS & METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Amir Z, Kwan SY, Landes D, Feber T, Williams SA (1999). Diagnostic delays in head and neck cancers. Eur J Cancer Care 8:198–203.

Ballard A, Green T, McCaa A, Logsdon MC (1997). A comparison of the level of hope in patients with newly diagnosed and recurrent cancer. Oncol Nurs Forum 24:899–904.[Medline]

Cheung YB (2000). Marital status and mortality in British women: a longitudinal study. Int J Epidemiol 29:93–99.[Abstract/Free Full Text]

Dimitroulis G, Reade P, Wiesenfeld D (1992). Referral patterns of patients with oral squamous cell carcinoma, Australia. Eur J Cancer B Oral Oncol 28B:23–27.

Elwood JM, Gallagher RP (1985). Factors influencing early diagnosis of cancer of the oral cavity. Can Med Assoc J 133:651–656.[Abstract]

Gorsky M, Dayan D (1995). Referral delay in diagnosis of oro/oropharyngeal cancer in Israel. Eur J Cancer B Oral Oncol 31(B):166–168.

Guggenheimer J, Verbin RS, Johnson JT, Horkowitz CA, Myers EN (1989). Factors delaying the diagnosis of oral and oropharyngeal carcinomas. Cancer 64:932–935.[Medline]

Hollows P, McAndrew PG, Perini MG (2000). Delays in the referral and treatment of oral squamous cell carcinoma [published erratum appears in Br Dent J 188(7):380, 2000]. Br Dent J 188:262–265.[Medline]

Johnson NJ, Backlund E, Sorlie PD, Loveless CA (2000). Marital status and mortality: the national longitudinal mortality study. Ann Epidemiol 10:224–238.[Medline]

Jovanovic A, Kostense PJ, Schulten EA, Snow GB, van der Waal I (1992). Delay in diagnosis of oral squamous cell carcinoma; a report from The Netherlands. Eur J Cancer B Oral Oncol 28(B):37–38.

Kantola S, Jokinen K, Hyrynkangas K, Mantyselka P, Alho OP (2001). Detection of tongue cancer in primary care. Br J Gen Pract 51:106–111.[Medline]

Kerawala CJ, Beale V, Reed M, Martin IC (2000). The role of vital tissue staining in the marginal control of oral squamous cell carcinoma. Int J Oral Maxillofac Surg 29:32–35.[Medline]

Kerdpon D, Sriplung H (2001). Factors related to delay in diagnosis of oral squamous cell carcinoma in southern Thailand. Oral Oncol 37:127–131.[Medline]

Kowalski LP, Franco EL, Torloni H, Fava AS, de Andrade Sobrinho J, Ramos G, et al. (1994). Lateness of diagnosis of oral and oropharyngeal carcinoma: factors related to the tumour, the patient and health professionals. Eur J Cancer B Oral Oncol 30(B):167–173.

Manski RJ (1995). Dental care coverage among older Americans. J Am Coll Dent 62:41–44.

Mashberg A, Merletti F, Boffetta P, Gandolfo S, Ozzello F, Fracchia F, et al. (1989). Appearance, site of occurrence, and physical and clinical characteristics of oral carcinoma in Torino, Italy. Cancer 63:2522–2527.[Medline]

Osterberg T, Lundgren M, Emilson CG, Sundh V, Birkhed D, Steen B (1998). Utilization of dental services in relation to socioeconomic and health factors in the middle-aged and elderly Swedish population. Acta Odontol Scand 56:41–47.[Medline]

Pisani P, Parkin DM, Bray F, Ferlay J (1999). Estimates of the worldwide mortality from 25 cancers in 1990 [published erratum appears in Int J Cancer 83:870-873, 1999]. Int J Cancer 83:18–29.[Medline]

Ries L, Eisner M, Kosary C, Hankey B, Miller B, Clegg L, et al. (2000). SEER cancer statistics review, 1973-1997. Bethesda, MD: National Cancer Institute.

Rodriguez Artalejo F, de Andres Manzano B, Guallar-Castillon P, Puente Mendizabal MT, Gonzalez Enriquez J, del Rey Calero J (2000). The association of tobacco and alcohol consumption with the use of health care services in Spain. Prev Med 31:554–561.[Medline]

Sciubba JJ (1999). Improving detection of precancerous and cancerous oral lesions. Computer-assisted analysis of the oral brush biopsy. US Collaborative OralCDx Study Group. J Am Dent Assoc 130:1445–1457.[Abstract/Free Full Text]

Silverman S Jr (1988). Early diagnosis of oral cancer. Cancer 62:1796–1799.[Medline]

Söderholm AL (1990). Oral carcinoma of the mandibular region. Br J Oral Maxillofac Surg 28:383–389.[Medline]

Wildt J, Bundgaard T, Bentzen SM (1995). Delay in the diagnosis of oral squamous cell carcinoma. Clin Otolaryngol 20:21–25.[Medline]

Yellowitz JA, Horowitz AM, Drury TF, Goodman HS (2000). Survey of US dentists' knowledge and opinions about oral pharyngeal cancer. J Am Dent Assoc 131:653–661.[Abstract/Free Full Text]

Zavras AI, Douglass CW, Joshipura K, Wu T, Laskaris G, Petridou E, et al. (2001). Smoking and alcohol in the etiology of oral cancer: gender-specific risk profiles in the south of Greece. Oral Oncol 37:28–35.[Medline]




This article has been cited by other articles:


Home page
 Alcohol AlcoholHome page
P. Barbadoro, D. Lucrezi, E. Prospero, and I. Annino
Improvement of Knowledge, Attitude, and Behavior About Oral Health in a Population of Alcohol Addicted Persons
Alcohol Alcohol., March 7, 2008; (2008) agn009v1.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (16)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Pitiphat, W.
Right arrow Articles by Zavras, A.I.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Pitiphat, W.
Right arrow Articles by Zavras, A.I.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
IADR Journals Advances in Dental Research ®
Journal of Dental Research ® Critical Reviews (1990-2004)