Introduction
Treatment for oral and oropharyngeal cancer can result in varying degrees of swallowing and speech dysfunction. Swallowing dysfunction is a multidimensional symptom complex in which patients can experience mechanical difficulties of bolus transfer, ineffective chewing and bolus formation, nasal regurgitation, oral leakage, and aspiration.Reference Wallace, Middleton and Cook 1 Speech dysfunction includes problems with speech production, intelligibility and acceptability of speech, and adequacy of speech in everyday situations.Reference Jacobi, van der Molen, Huiskens, van Rossum and Hilgers 2 Both speech and swallowing dysfunction impact significantly on aspects of health-related quality of life (QoL),Reference Chen, Frankowski, Bishop-Leone, Hebert, Leyk and Lewin 3 , Reference Rinkel, Verdonck de Leeuw, Langendijk, van Reij, Aaronson and Leemans 4 including emotions, self-esteem and social functioning. Consequently, speech and swallowing dysfunction is a significant issue for oral and laryngeal cancer survivors.
Although post-treatment swallowing dysfunction is relatively common, it remains under-reported in the oral and oropharyngeal cancer population.Reference Dwivedi, St Rose, Roe, Khan, Pepper and Nutting 5 Similarly, the extent of post-treatment speech problems in this cohort is inadequately quantified, due to a lack of uniformity in the assessment and measurement of speech outcomes.Reference Dwivedi, Kazi, Agrawal, Nutting, Clarke and Kerawala 6 In this respect, the role of patient-reported post-treatment swallowing and speech outcomes is becoming increasingly important, because of the shift toward supported self-management amongst cancer survivors. 7
Speech outcomes are predominantly evaluated by patient-reported questionnaires, of which the University of Washington Quality of Life questionnaireReference Hassan and Weymuller 8 is the most frequently used.Reference Dwivedi, Kazi, Agrawal, Nutting, Clarke and Kerawala 6 Objective assessments of dysphagia have inherent barriers to routine use in the out-patient setting, as they often require additional hospital appointments to undertake the multiple tests frequently employed to reach a definitive diagnosis. Patient-reported swallowing outcome measures have been used extensively to assess post-treatment dysphagia, utilising instruments such as the M D Anderson Dysphagia Inventory,Reference Chen, Frankowski, Bishop-Leone, Hebert, Leyk and Lewin 3 the Swallowing Quality of Life InstrumentReference Rinkel, Verdonck de Leeuw, Langendijk, van Reij, Aaronson and Leemans 4 and the Sydney Swallow Questionnaire.Reference Dwivedi, St Rose, Roe, Khan, Pepper and Nutting 5 While these questionnaires can provide detailed self-reported evaluation of swallowing function, there is currently no single questionnaire that provides a quick, simple, self-reported screening tool assessing dysfunction and/or the need for supportive care, and which can be utilised routinely in the clinic.
Routine screening in the out-patient setting may potentially reduce the prevalence of unrecognised swallowing and speech dysfunction, and can facilitate instigation of supportive care for this problem. Instruments intended to measure health-related QoL in head and neck cancer patients have in the past also been utilised as self-reported swallowing and speech dysfunction screening tools, because the majority of these tools frequently assess swallowing and speech within the functional domain.Reference Rogers and Lowe 9
However, if a routinely administered health-related QoL questionnaire could collect health-related QoL data while also screening for dysfunction, this would reduce the burden of multiple questionnaires imposed upon the patient.
The University of Washington QoL questionnaire employs sophisticated methodology to enable the differentiation of QoL impairment on the basis of severity,Reference Rogers and Lowe 9 by setting questionnaire score cut-off thresholds (beyond which significant problems are indicated) for domains including speech and swallowing. Subsequent studies have found a good correlation between physiological investigation findings and University of Washington QoL questionnaire patient-reported speech and swallowing outcomes, suggesting that this questionnaire is applicable as a quick screening tool.Reference Thomas, Jones, Tandon, Katre, Lowe and Rogers 10 , Reference Thomas, Jones, Tandon, Carding, Lowe and Rogers 11
However, health-related QoL tools are not holistic in nature, because of their rigid construction based on exacting psychometric specifications. Furthermore, the potential modification of such tools to enable their use as dysfunction screening instruments (in order to assess the patient's need for supportive care) could invalidate the tool's ability to measure QoL, hindering longitudinal, multicentre comparisons.
The Patient Concerns Inventory is a holistic, patient-reported tool designed to screen for unmet needs by allowing patients to identify (prior to their out-patient specialist consultation) concerns that they wish to discuss during their consultation.Reference Rogers, El-Sheikha and Lowe 12 The Patient Concerns Inventory also allows patients to choose, from a list of healthcare professionals, those with whom they wish to speak. Clinicians who use the Patient Concerns Inventory are more likely to address their patients' unmet needs, as they are able to focus on each patient's specific concerns, thus facilitating appropriate and timely supportive care interventions.Reference Ghazali, Kanatas, Langley, Scott, Lowe and Rogers 13
This study aimed to assess the value of the Patient Concerns Inventory for routine screening for self-reported swallowing and speech concerns in the out-patient setting. The study also aimed to evaluate the Inventory's utility in assessing patients' supportive needs (in relation to their speech or swallowing impairment) from a speech and language therapy perspective, in order to determine whether the Inventory had additional clinical value in this respect.
Method
The patients recruited to this study comprised one consultant's (SNR's) oral and oropharyngeal cancer patients attending out-patient clinics from 1 August 2007 to 15 July 2009. Patients were included if they were disease-free and receiving routine follow up. Patients were excluded if they were pre-treatment, receiving palliative care, attending the dressing clinic for post-operative wound management, or already part of another outcomes study.
Study data were collected using self-administered surveys delivered via a touch-screen computer.Reference Millsopp, Frackleton, Lowe and Rogers 14 Patients used a Microsoft Access based software package which delivered the University of Washington QoL questionnaire together with the Patient Concerns Inventory. Survey answer data were transferred directly onto the hospital computer drive. Using normal password protection arrangements, these data were retrieved by the consultant in the consultation room immediately before seeing the patient.
The Patient Concerns Inventory comprises a list of 15 professionals with whom patients may wish to speak, together with a check-list of 54 items of potential patient concern (in our study, 45 items were used from August 2007, with a further nine items added in April 2008).Reference Rogers, El-Sheikha and Lowe 12 , Reference Millsopp, Frackleton, Lowe and Rogers 14 The Patient Concerns Inventory items relevant to speech and swallowing are ‘speech/voice being understood’, ‘swallowing’ and (from April 2008) ‘regurgitation’; the relevant professionals are ‘dietician’ and ‘speech and language therapist’.
The study also used the University of Washington QoL questionnaire version four,Reference Rogers, Gwane, Lowe, Humphris, Yueh and Weymuller 15 a standard, cancer-specific head and neck questionnaire with 12 domains, including swallowing and speech. These domains are scaled from 0 (worst) to 100 (best) according to the hierarchy of response. The swallowing domain has four response options: 0 = ‘I cannot swallow because it goes down the wrong way and chokes me’, 30 = ‘I can only swallow liquid food’, 70 = ‘I cannot swallow certain solid foods’ and 100 = ‘I can swallow as well as ever’. The speech domain also has four response options: 0 = ‘I cannot be understood’, 30 = ‘only my family and friends can understand me’, 70 = ‘I have difficulty saying some words but I can be understood over the phone’ and 100 = ‘my speech is the same as always’. Based on earlier work,Reference Thomas, Jones, Tandon, Katre, Lowe and Rogers 10 we defined significant swallowing or speech problems as a University of Washington QoL questionnaire score of 0 or 30 in either domain.
Throughout the rest of this paper, we refer to patients with University of Washington QoL questionnaire swallowing or speech domain scores of 0 or 30 as having ‘significant swallowing or speech problems’. Patients who did not meet these University of Washington QoL criteria but whose Patient Concerns Inventory responses indicated speech or swallowing concerns (i.e. patients wanting to discuss relevant issues or wanting to talk with a dietician or speech and language therapist) are referred to as having ‘more minor swallowing or speech concerns’.
For the purpose of Patient Concerns Inventory data collection, new onward referral data were obtained either retrospectively via scrutiny of clinic letters (up to 24 September 2008) or prospectively via direct observation of the consultation (undertaken by BS) (from 25 September 2008 to 15 July 2009).
To enable comparison, we also analysed University of Washington QoL questionnaire swallowing and speech data from 372 non-cancer patients attending 10 general dental practices.Reference Rogers, O'Donnell, Williams-Hewitt, Christensen and Lowe 16
We compiled a list of patients with swallowing or speech issues, as indicated by the University of Washington QoL questionnaire and the Patient Concerns Inventory. This list of patients was then reviewed by the speech and language therapist to determine whether individual patients were already known to their service and/or were receiving treatment.
Clinical and demographic data were obtained from the Liverpool head and neck cancer database.
Since clinic appointment data were not independent of patient data, and since patients differed in the number of consultations they attended, any p values computed were adjusted for possible patient clustering effects using Stata binary regression software. A p value of less than 0.01 was regarded as statistically significant, to compensate for the large number of statistical tests performed.
Research ethics approval for this study was issued by the Sefton research ethics committee.
Results
During the study period, 204 patients completed the two computerised surveys, at 454 clinic appointments. The mean (standard deviation) age of patients completing the surveys at their consultation was 62 (11) years, and these consultations occurred at a median (interquartile range) time of 18 (eight to 47) months after the patient's cancer diagnosis. Of the 204 patients, 57 per cent (117) were male and 43 per cent (87) female, 78 per cent (160) had a squamous cell carcinoma, 68 per cent (139) had oral cancer, 78 per cent (144/186) had cT1 to T2 tumours, 19 per cent (35/188) had cN(+) tumours, 44 per cent (90) had received head and/or neck radiotherapy at some time since diagnosis, and 48 per cent (98) had undergone free-flap surgery.
Of the clinic appointments for which patients completed the University of Washington QoL questionnaire prior to consultation, a significant swallowing problem (i.e. a score of 0 or 30) was reported for 21 per cent (95/454) and a significant speech problem (i.e. scores of 0 or 30) for 7 per cent (34/454). One or the other problem was reported for 22 per cent of all consultations (101/454).
In contrast, of the non-cancer patients attending general dental practice appointments, only 2.2 per cent (eight of 372) had significant swallowing problems, 0.8 per cent (three of 372) had significant speech problems and 2.4 per cent (nine of 372) had either problem.Reference Rogers, O'Donnell, Williams-Hewitt, Christensen and Lowe 16
Of the clinic appointments for which patients completed the Patient Concerns Inventory prior to consultation, swallowing was raised as a discussion issue for 17 per cent (79/454) of consultations and speech as a discussion issue for 13 per cent (57/454). In a similar fashion, of the consultations occurring after April 2008, regurgitation issues were raised for 1 per cent (three of 267 consultations); in two of these three consultations, swallowing was also an issue.
The Patient Concerns Inventory was completed for 454 consultations; for 4 per cent (16/454) of these, the patient requested to speak to a dietician, and for 7 per cent (33/454) they requested to speak to a speech and language therapist (either in the clinic itself or via referral). Overall, the Patient Concerns Inventory noted a swallowing, speech and/or regurgitation concern, or a request to speak to a dietician or a speech and language therapist, for 28 per cent (128/454) of consultations.
Considering both surveys together, 39 per cent (178/454) of consultations (involving 48 per cent (97/204) of patients) raised significant swallowing or speech problems (as per the University of Washington QoL questionnaire) or highlighted more minor swallowing or speech concerns (as per the Patient Concerns Inventory) (Table I). Of these 178 consultations, 101 involved patients who met the University of Washington QoL questionnaire criteria for significant swallowing or speech problems, while 77 involved patients who met only the Patient Concerns Inventory criteria for more minor swallowing or speech concerns. In half (50/101) of all consultations involving a significant swallowing or speech problem, the patient did not wish to discuss the problem.
Table I Swallowing and speech responses: university of washington quality of life questionnaire vs patient concerns inventory
Patient Concerns Inventory (PCI) data represent number of consultations. See text for University of Washington Quality of Life questionnaire (UWQoL) swallowing and speech domain scoring systems. *Consultations for which patients had significant swallowing or speech problems (on UW-QoL) or more minor swallowing or speech concerns (on PCI).
Significant swallowing or speech problems (as per the University of Washington QoL questionnaire) were reported more frequently by patients with more advanced tumours, by those who had received head or neck radiotherapy, and by those who had undergone free-flap surgery (Table II).
Table II Patient factors vs swallowing and speech responses
*Comparing consultations for which significant swallowing or speech problems (SSP) were noted on the University of Washington Quality of Life questionnaire (UWQoL) (n = 101), versus those for which such problems were not noted (n = 353), adjusted for patient clustering effects using binary regression (see Methods). †Numerical values for age and time since diagnosis (Dx) were used in the regressions. ‡Oral vs pharyngeal. **Denominator excludes. Tx = Primary tumour cannot be assessed, Nx = Regional lymph nodes cannot be evaluated and unknown staging. PCI SSC = Patient Concerns Inventory swallowing and speech concerns (no SSP on UWQoL); Neither = neither UWQoL SSP nor PCI SSC; SCC = squamous cell carcinoma; T = tumour stage; H&N RT = head and neck radiotherapy; FF Sx = free-flap surgery
Patients with significant swallowing or speech problems and those with more minor swallowing or speech concerns differed little in their responses to the non swallowing and speech items on the Patient Concerns Inventory. One notable exception was the percutaneous endoscopic gastrostomy tube item: almost all the patients who selected this item had significant swallowing or speech problems. However, the combined group comprising patients with significant swallowing or speech problems and patients with more minor swallowing and speech concerns was much more likely to want to discuss almost every other PCI item (i.e. other than percutaneous endoscopic gastrostomy tube), compared with patients without swallowing or speech issues (Table III). An exception to this was the fear of cancer recurrence item: the proportion of consultations in which patients expressed a wish to discuss this fear ranged from 30 per cent for patients without swallowing or speech issues, through 35 per cent for those with significant swallowing or speech problems, to 43 per cent for those with more minor swallowing or speech concerns.
Table III Patient concerns inventory items raised, by swallowing and speech response
Total number of consults = *101, †77 and ‡276. **Adjusted for patient clustering effects using binary regression (see Methods). §Diarrhoea or constipation. #Smoking or alcohol. ¥Total number of consults from April 2008 to end of study (used to calculate group percentages for new items): 58 for group 1; 44 for group 2; 165 for group 3. PCI = Patients Concerns Inventory; Grp = group; UWQoL SSP = University of Washington Quality of Life questionnaire significant swallowing or speech problems; SCC = more minor swallowing or speech concerns; Neither = neither UWQoL SSP nor PCI SSC; DN = district nurse; PEG = percutaneous endoscopic gastrostomy; NC = not calculable
The type of health professional with whom patients wanted to speak did not differ much, comparing patients with significant swallowing or speech problems versus those with more minor swallowing or speech concerns (data not shown). These two groups combined were more likely to want to talk with a dentist, dietician, oral rehabilitation team member, speech and language therapist, and/or family doctor, compared with patients without swallowing or speech issues. There were no obvious differences with regard to new onward referrals, except that the combined group was more often referred to the oral rehabilitation team and to the speech and language therapist, compared with patients without swallowing or speech issues (data not shown).
Of the 97 patients who (in 178 consultations) had either significant swallowing or speech problems or more minor swallowing or speech concerns, 23 (24 per cent) were unknown to the speech and language therapist.
Of these 23 patients, only three required onward referral. One was an 80-year-old man with stage one tongue cancer, who had both speech and swallowing problems. The second was a 46-year-old woman with stage two tongue cancer, who had speech problems. The third was a 67-year-old patient with stage four disease and swallowing problems. Speech was assessed using intelligibility scores and swallowing by videofluoroscopy. All three patients received speech and language therapy aimed at improving function, including training in compensatory techniques.
Discussion
Our study findings indicated that the Patient Concerns Inventory was a valuable tool for routine screening of self-reported swallowing and speech dysfunction, and enabled patients' concerns to be addressed during out-patient clinic consultations. The Patient Concerns Inventory was also a valuable patient-reported tool with which to identify patients requiring supportive care, as it identified those patients who were already receiving specialist care for perceived swallowing and/or speech difficulties, as well as those who required such care. Thus, the Patient Concerns Inventory appears to have the potential to facilitate multidisciplinary, holistic, patient-centred care for oral and oropharyngeal cancer patients.
This study may have been limited by inclusion of a selective study population consisting of patients seen by a single surgical consultant, who had integrated the Patient Concerns Inventory into his routine clinical practice. Recruitment of consecutive clinic attendees may have reduced some of this selection bias. While this study did not use any objective measure of swallowing or speech to determine the severity of the problem, the University of Washington questionnaire is known to compare favourably with Both, objective and validated patient-reported measures of swallowing and speech.Reference Thomas, Jones, Tandon, Katre, Lowe and Rogers 10 , Reference Thomas, Jones, Tandon, Carding, Lowe and Rogers 11
By allowing patients the opportunity to discuss, within the consultation, their concerns as identified by the Patient Concerns Inventory, the clinician can be more focused and appropriate in managing those concerns. In our study, the clinician was able to deal with many patient-reported swallowing or speech concerns in the clinic, reserving onward referral (based on objective clinical judgement) for a much smaller number of patients. Post-treatment speech and/or swallowing dysfunction is common, and the fact that some of our patients appeared reluctant to broach the subject during the consultation emphasises the clinician's responsibility to initiate discussion on this issue. In our study, 50 per cent of patients with significant swallowing or speech problems identified on the University of Washington QoL questionnaire did not highlight these as issues for discussion when completing the Patient Concerns Inventory (despite technically successful completion of this survey). In some cases, this may have been due to the patient being a cancer survivor who had accepted and adapted to their deficit; however, other patients may have had communication barriers (e.g. a reluctance to complain, low self-esteem, or the assumption that the information was not important as the clinician did not enquire about it).
Patients who self-reported significant swallowing or speech problems on the University of Washington QoL questionnaire were more likely to have more advanced tumours, to have received head or neck radiotherapy, and to have undergone free-flap surgery. This echoes the findings of another study of swallowing and speech outcomes.Reference Dwivedi, St Rose, Roe, Khan, Pepper and Nutting 5
Patients with prolonged dysphagia may need an alternative means of feeding, e.g. a percutaneous endoscopic gastrostomy tube, in order to ensure their nutritional requirements are met. However, such tubes can have complications;Reference Paleri and Patterson 17 furthermore, they are often poorly accepted by patients due mainly to their impact on appearance and intimacy.Reference Rogers, Thomson, O'Toole and Lowe 18 Unsurprisingly, those patients who raised percutaneous endoscopic gastrostomy tubes as an item of concern on the Patient Concern Inventory almost invariably had significant swallowing or speech problems on the University of Washington QoL questionnaire. Patients are often extremely motivated swiftly to regain an acceptable swallowing function, in order to have such tubes removed.
The other Patient Concerns Inventory item frequently highlighted by all patients, including those with significant swallowing or speech problems, was fear of cancer recurrence. It is known that some patients perceive the presence of a symptom or problem as an indication of cancer recurrence,Reference Rogers, Scott, Lowe, Ozakinci and Humphris 19 which can trigger fear of a recurrence along with general fear of the unknown.Reference Lee-Jones, Humphris, Dixon and Hatcher 20 Patients with on-going problems requiring repeated investigations may also feel anxious in case their clinician has missed a recurrence.Reference Lee-Jones, Humphris, Dixon and Hatcher 20 Head and neck cancer survivors suffer much fear of recurrence. Addressing this anxiety in general, and specifically in patients with swallowing or speech problems, can help reduce the patient's associated psychological distress.Reference Humphris and Ozakinci 21
The process of screening patients for unmet support needs requires identification of areas in which they may require help in order to maintain or regain optimal health and QoL.Reference Foot and Sanson-Fisher 22 Such screening forms an important part of the national cancer survivorship agenda. 7 Our study findings suggest that patients can accurately identify their speech and swallowing dysfunction. Head and neck cancer survivors are capable of accurately detecting changes in their swallowing function,Reference Pauloski, Rademaker, Logemann, Lazarus, Newman and Hamner 23 and such self-reporting can be a valuable source of feedback over the course of any therapeutic intervention.
In our study, most of the patients who believed they had swallowing or speech issues (as assessed by the combined Patient Concerns Inventory and University of Washington QoL questionnaire surveys) were already known to our speech and language therapy service (77 per cent; 74/97 patients), having been referred at the time of surgery. These patients' Patient Concerns Inventory data verified that they were under speech and language therapist care (i.e. receiving active treatment or under review). Those patients who were not currently receiving speech and language therapy intervention had reached a plateau in their speech and swallowing recovery. In such cases, problems with adaptation may exist, and discussing these issues may identify the need for other services (e.g. psychological support).
In the three patients whose Patient Concerns Inventory results generated a new speech and language therapy referral, assessment confirmed the presence of impairment, and appropriate intervention was carried out (data not shown). In all three of these patients, the use of compensatory techniques and direct therapy was found to be effective. Outcomes were not formally evaluated, but there were improvements in patient- and clinician-rated intelligibility and swallowing function. In future studies, further evaluation using disorder-specific questionnaires and therapy outcome measures would be valuable.
Not infrequently, some patients' support needs are overlooked by the multidisciplinary head and neck oncology team, due to logistical and/or administrative problems, and this may lead to an accumulation of unmet needs. In this situation, the use of health-related QoL questionnaires alone will not detect the problem efficiently. However, routine use of the Patient Concerns Inventory in the clinic may enable patients to raise their support needs in an effective manner.
In our study, the combined use of the University of Washington QoL questionnaire and the Patient Concerns Inventory highlighted 23 patients with significant swallowing or speech issues who were unknown to the speech and language therapist. The majority of these patients had their concerns dealt with in the clinic by the consultant, leaving only three patients requiring an onward referral. Of these three patients, one had received primary treatment abroad but had returned for follow up at our hospital, and had thus missed the customary pre-treatment speech and language therapist evaluation. The other two patients received primary treatment at our hospital but were not referred to the speech and language therapist during their in-patient stay, because of a relatively short admission and/or the perception that early post-operative problems appeared to be resolving at discharge. Use of the Patient Concerns Inventory during review consultations helped these patients to raise unresolved problems with the oncology team, and empowered them to be proactive in their care.
-
• Speech and swallowing dysfunction is common in head and neck cancer but under-reported
-
• This study assessed this dysfunction using the University of Washington Quality of Life questionnaire, which correlates well with objective measures
-
• It also used the Patients Concerns Inventory, a patient-reported needs assessment tool
-
• This Inventory identified previously undetected swallowing and/or speech dysfunction
-
• Use of both surveys identified patients with significant swallowing or speech issues who did not want to discuss them
-
• Clinicians should enquire about such issues during consultations
The issue of access to speech and language therapy services is crucial, as not all multidisciplinary oncology centres have the same level of access to these services, despite the fact that such therapists are core members of the multidisciplinary oncology team. 24 Thus, the Patient Concerns Inventory can provide a ‘safety net’ to identify patients who would otherwise fail to receive appropriate multidisciplinary supportive care.
In our study, patients who had either significant swallowing or speech problems (as per the University of Washington QoL questionnaire) or more minor swallowing or speech concerns (as per the Patient Concerns Inventory) were more likely to want to see a dentist, dietician, oral rehabilitation team member, speech and language therapist, and/or family doctor, compared with patients without swallowing or speech issues. This list of health professionals reflects the multidimensional nature of swallowing and speech dysfunction, requiring the input of various professionals for supportive care. Nevertheless, onward referrals were more often made to speech and language therapists and the oral rehabilitation team. Theoretically, the multidisciplinary nature of oncology clinics should allow patients to be seen by other specialists on the same day as their clinic appointment; however, onward referrals may still be required on those occasions when individual team members are not available on a particular day, or when further investigation is required in a different department.
Conclusion
Speech and swallowing dysfunction is common but is under-reported in head and neck cancer survivors. Patient-reported screening tools to detect such dysfunction, based on health-related QoL questionnaires, have been shown to correlate well with objective measures.
The Patient Concerns Inventory is a holistic, patient-reported needs assessment tool suitable for use in the out-patient setting. In this study, this Inventory was able to accurately identify patients with swallowing and/or speech dysfunction, both those already known to the speech and language therapy service and those with an unmet need for this service.
This study found that concurrent use of the Patient Concerns Inventory and the University of Washington QoL questionnaire identified a number of patients with significant swallowing or speech issues but without the desire to discuss these issues with their consultant. Thus, we believe that clinicians should take the initiative and broach these issues during the consultation.
