Frequency of office visits for groups 1, 2, and 6, with upper and lower limits of recommended guidelines.6 Only stage II data are shown (stages I and III/IV were similar). SHNS indicates Society of Head and Neck Surgery; ASHNS, American Society for Head and Neck Surgery.
Frequency of office visits for all respondents (group 6), given as percentage practicing within or varying from Task Force guidelines.6 Data shown are the mean usage rates of the 3 stage groupings. The degrees of overuse and underuse both increase with advancing postoperative year.
Use of chest radiographs by American Society for Head and Neck Surgery respondents (group 4), given as percentage practicing within or varying from Task Force guidelines.6 Data shown are for stage II only (stages I and III/IV were similar). Overuse decreases and underuse generally increases with advancing postoperative year.
Use of liver function tests by American Society for Head and Neck Surgery respondents (group 4), given as percentage practicing within or varying from Task Force guidelines.6 Data shown are for stage III/IV only (stages I and II were similar). Overuse decreases and underuse increases with advancing postoperative year.
Paniello RC, Virgo KS, Johnson MH, Clemente MF, Johnson FE. Practice Patterns and Clinical Guidelines for Posttreatment Follow-up of Head and Neck CancersA Comparison of 2 Professional Societies. Arch Otolaryngol Head Neck Surg. 1999;125(3):309-313. doi:10.1001/archotol.125.3.309
To determine and compare the current follow-up practice patterns of members of 2 professional societies of head and neck surgeons, and to compare these with the societies' published clinical practice guidelines.
A survey was mailed to the 640 members of the American Society for Head and Neck Surgery (ASHNS); results were compared with those of a similar survey of the 824 members of the Society of Head and Neck Surgery (SHNS) and with the clinical practice guidelines of the consensus committee of both societies.
Main Outcome Measures
Data were collected regarding the frequency of follow-up visits after potentially curative resection of head and neck epidermoid carcinoma and the types of diagnostic studies performed at each visit.
A total of 318 ASHNS members responded to 1 of the mailings (49.7%), of which 280 responses (43.8%) were evaluable. Most surgeons relied on directed history, physical examination, and routine chest radiograph at varying intervals for detection of recurrences and second primary tumors. Other tests were used sporadically. For frequency of follow-up testing, the percentage of surgeons who followed the published guidelines varied from 97% in postoperative year 1 to 62% in postoperative year 5. A mean of 24% of surgeons varied from the guidelines in their use of chest radiographs, and 45% varied in their use of liver function tests. The ASHNS members used significantly more office visits than the SHNS members during the first 2 postoperative years.
The strategies used by members of the ASHNS and the SHNS for posttreatment surveillance after potentially curative resection of malignant neoplasms of the head and neck were generally similar but showed some important differences. Most surgeons used directed history and physical examination at regular intervals, and annual chest radiographs. The follow-up practices of most members of these societies, which have recently merged, fall within the recommendations of the Clinical Guidelines Task Force.
SURVEILLANCE strategies after potentially curative resection of head and neck cancers typically aim to detect local and regional recurrence of the primary disease, distant metastases, and second primary tumors. No single strategy has been generally agreed on, but most surgeons believe that some routine follow-up is important.1 Boysen et al2,3 reported that 61% of recurrences were found after patients self-referred for a specific symptom, rather than on scheduled routine follow-up, but deVisscher and Manni4 found that the mean survival of patients with recurrences detected on routine follow-up was nearly double that of patients detected after self-referral. This group also pointed out the psychological value of reassuring patients of normal results of a follow-up examination.
Recommendations for follow-up strategies have received little attention in the literature or in major textbooks.5 In 1996, the American Society for Head and Neck Surgery (ASHNS) and the Society of Head and Neck Surgery (SHNS) collaborated to form a Clinical Guidelines Task Force, resulting in a booklet of consensus guidelines.6 These include postoperative follow-up recommendations, given by site and stage for each major type of head and neck cancer. The recommendations are summarized in Table 1.
We previously reported the results of a survey of the membership of the SHNS.1 Marchant et al7 performed a similar survey of the members of the ASHNS in 1992. The purposes of the present study were to describe the current follow-up strategies of ASHNS members by means of our survey instrument, to compare the 2 societies, and to determine how closely the responding surgeons follow the recommendations of the Clinical Guidelines Task Force. In October 1997, after the data collection from these surveys was completed, the 2 societies agreed to merge into a new organization, the American Head and Neck Society, making examination of the pooled data from the 2 surveys particularly appropriate.
Membership lists and permission to perform the surveys were obtained from the leadership of both societies. The 3-page surveys were mailed in September and December 1996. All members of both societies, domestic and international, were surveyed, and nonresponders were sent a second survey 1 month later. There were 824 members of the SHNS and 640 members of the ASHNS on the lists; however, a comparison disclosed that 118 surgeons were members of both societies. Care was taken to ensure that this overlap group was surveyed only once. Member surgeons who did not currently perform head and neck cancer surgery or who did not follow up their own patients postoperatively were asked to indicate such and return the survey uncompleted. The remaining surgeons were asked to indicate how frequently they performed office examination (with or without laryngoscopy) and 13 other examinations and blood tests.
The modalities to be included in the survey were determined by a review of the literature and by discussion with local members of each society. The imaging modalities included chest radiography, computed tomography of the head and neck, computed tomography of the chest, magnetic resonance imaging of the head and neck, head and neck sonogram, and bone scan. The blood tests included complete blood count, electrolytes (with or without calcium), liver function tests, thyroid function tests, and specific tumor markers. Other examinations included bronchoscopy and esophagoscopy.
Frequency of use was expressed as the number of times a test was performed annually; thus, if an examination was performed every 2 months, the frequency was 6 times per year. The frequency data are reported as the mean of all evaluable responses. Data were stratified for each of the first 5 years after potentially curative treatment for head and neck epidermoid carcinoma, and were further divided into 3 TNM stage groups (I, II, and III plus resectable IV). The survey instrument has been published previously in its entirety.1 The completed surveys were entered into a computerized database and analyzed. Statistical software was used to perform analysis of variance and z tests with the level of statistical significance set at P<.05.
A total of 318 (49.7%) of the 640 ASHNS members returned 1 of the surveys. There were 30 members who reported they did not currently perform head and neck cancer surgery, 5 who did not conduct postoperative follow-up, and 3 who were retired, leaving 280 evaluable responses (43.8%). The SHNS survey had a 43.9% response rate and a 24.9% evaluable rate.1 The presence of the large number of members of both societies resulted in 6 comparison groups (Table 2). The 59 evaluable responses from surgeons who were members of both groups (group 3) represent 21.1% of the 280 total evaluable ASHNS responses (group 4) and 29.6% of the 199 evaluable all-SHNS responses (group 5). Unless otherwise indicated, references to the ASHNS and SHNS groups will denote groups 4 and 5, respectively (ie, the dual member data will usually be included in the analysis).
The demographic distribution of the ASHNS respondents is shown in Table 3. The mean age was about 50 years. There were more respondents from private practice (53.9%) than from academic practice (41.1%). Only 26.1% of the respondents reported spending more than half of their time treating patients with head and neck cancer. These demographic data were similar to those of the SHNS group, except that the ASHNS group had a greater proportion of respondents in private practice than the SHNS group did.
The frequency of follow-up office visits is shown by stage for all survey groups in Table 4. The data for groups 1, 2, and 6, along with the range of the society recommendations, are shown graphically for stage II in Figure 1. The mean survey values fell well within the recommended guideline range, except for postoperative year 5, when all survey group means dipped just below the suggested 2 visits per year. Follow-up intensity for the ASHNS-only group (group 1) differed significantly from that of the SHNS-only group (group 2) during years 1 and 2 (z test, P<.05), but not for later years. When the dual member data (group 3) were added to each group, this difference (group 4 vs group 5) remained statistically significant for the first 2 postoperative years. Follow-up frequency for the total group (group 6) was significantly different from the SHNS-only group (group 2) for the first 2 years and from the all-SHNS group (group 5) for the first year only. None of the other comparisons between study groups showed any statistically significant differences for frequency of office visits.
When the frequency of postoperative office visits for the total group (group 6) was compared with the consensus recommendations of the 2 societies, most fell within the recommended range, but some did not; these are summarized in Figure 2. A large group of surgeons reported following their patients only once yearly in postoperative years 4 and 5, while the recommendation is for continued follow-up at 4- to 6-month intervals (Table 1). A few surgeons also followed their patients at least every 2 months during all 5 postoperative years.
For each group, there was a statistically significant increase in the frequency of visits with increasing TNM stage (Table 4). In the ASHNS survey group (group 4), 70.6% reported that they used the same follow-up regimen for stages II, III, and IV, while 29.4% did not. The Task Force recommendations are the same for tumor stages I/II as for stages III/IV.
The societies recommend annual chest radiography. For the ASHNS group (group 4), there was a statistically significantly downward trend in the frequency of performing chest radiography with advancing postoperative years, decreasing from a mean of 1.18 times during the first year to 0.85 time in year 5. There was also a small but significant increase in use with advancing tumor stage. About one fourth of respondents varied from the Task Force recommendations (Figure 3), often ordering more than 1 chest radiograph in each of the first 2 postoperative years, and less than 1 in the final 3 years. These results were similar for each of the other 5 study groups, with no statistically significant differences identified.
The other recommendation is for annual liver function tests. A majority of ASHNS members ordered this test less than once per year, while about 10% used it more than once (Figure 4); the mean frequency in postoperative year 1 ranged from 0.55 time for stage I to 0.73 time for stage III/IV. The SHNS group ordered liver function tests 65% to 75% more often than the ASHNS group for all years and all stages; for the first year, the frequency ranged from 0.97 time for stage I to 1.23 times for stage III/IV. The higher frequency came primarily from some SHNS members who ordered more than 1 set of liver function tests each postoperative year; 35% to 50% of members did not use liver function tests at all, similar to the ASHNS group. The differences between the per-stage, per-year mean frequencies were statistically significant at P<.05.
The clinical guidelines recommend annual thyroid function tests for patients who had radiation therapy delivered low in the neck. Since the survey did not specify this variable, it cannot be determined how closely this recommendation was followed. Nearly half (47%) of the ASHNS group reported using thyroid function tests at some point in their routine postoperative follow-up, with a mean of 0.74 time per year for stage III/IV, year 1 (the frequency was lower for other years and stages). The other subgroups had similar usage rates with no statistically significant differences.
Other tests and examinations in the survey were used by some of the respondents, although none is recommended for routine use by the Task Force. Complete blood count, electrolytes, and head and neck computed tomographic scan were used by more than one fourth of respondents at some point in their routine follow-up. The frequency of use was highest in the first postoperative year, declining in subsequent years. There was also a significant increase in use with increasing tumor stage, as with the other examinations studied. The first-year data for the ASHNS group, averaged across tumor stage, are given in Table 5. The SHNS group had similar rates of usage, with no statistically significant differences.
This is the first study, to our knowledge, to compare the behaviors of head and neck surgeons who are members of the 2 major head and neck societies. The majority of members of both societies practice within the clinical practice guidelines recommended by the joint Task Force. The study shows some statistically significant differences among the respondents, but these differences are small and are probably not clinically important; rather, they probably reflect differences in philosophy of the various training programs. Still, the 2 societies are clearly more similar than they are different in terms of postoperative follow-up management. This finding is pertinent because of the recent merger of the 2 societies into the American Head and Neck Society.
An interesting group is group 3, the respondents who are members of both societies. The behavior of this group resembled that of the ASHNS group more than the SHNS group, and as a result the practice patterns of groups that included these members (groups 3, 4, and 6) resembled the ASHNS-only group (group 1) more than the SHNS-only group (group 2).
The surveys indicate that members of both societies believe that postoperative follow-up is important, and that they perform these examinations themselves rather than leaving the follow-up to another physician. The response rate of nearly 50% is reasonable for surveys of this type. Assuming that the respondents' practice patterns are similar to those of the nonrespondents, these results are likely to represent the opinions of most experts in the field of head and neck cancer.
The implications of this study's major findings may be most clearly understood in economic terms. A detailed analysis of the financial implications of various follow-up strategies is presented in a recently published study.5 Given the costs of all medical tests and procedures, it will likely become increasingly difficult to justify routine use of follow-up strategies that are more intense than the recommendations of the Clinical Guidelines Task Force. The guidelines wisely leave room for case-specific modifications. Given the absence of well-controlled studies demonstrating the superiority of any specific follow-up strategy over alternative strategies, the guidelines provide a valuable reference and appear to fairly represent the current practice of most North American head and neck surgeons.
The members of the ASHNS and the SHNS use similar strategies for surveillance after potentially curative resection of malignant neoplasms of the head and neck. These include directed history and physical examination at regular intervals and annual chest radiographs. The follow-up practices of most members of these societies, which have recently merged, fall within the recommendations of the Clinical Guidelines Task Force.
Accepted for publication October 13, 1998.
Presented at the spring meeting of the American Head and Neck Society, Palm Desert, Calif, April 25, 1999.
Reprints: Randal C. Paniello, MD, 517 S Euclid, Box 8115, St Louis, MO 63110 (e-mail: email@example.com).