SEARCH

SEARCH BY CITATION

Keywords:

  • Oral Health Impact Profile;
  • oral health-related quality of life;
  • sensitivity to change;
  • Short Form-36;
  • treatment effects

Abstract

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. Conflicts of interest
  8. References

The literature presents conflicting findings on whether health-related quality of life (HRQoL) measures have sufficient evaluative properties to assess changes caused by dental interventions. The aim of our study was to compare sensitivity to change in HRQoL and OHRQoL in prosthodontic patients. In this prospective intervention study, a total of 165 consecutively recruited patients completed the Short Form-36 (SF-36) and the 49-item Oral Health Impact Profile (OHIP), as self-administered questionnaires, before prosthodontic treatment and 1 month after treatment was finished. Differences in SF-36 and OHIP scores between baseline and follow up were tested for statistical significance using paired t-tests. Effect sizes (Cohen's d) were calculated. Health-related quality of life improved during prosthodontic treatment, indicated by a slight, but statistically significant, increase in the SF-36 physical component (difference: 1.0 points), whereas perceived mental health did not change substantially (difference: −0.5 points). Improvement in OHRQoL (difference in OHIP sum score: −6.7 points) was statistically significant. Although the OHIP effect size (of 0.2) was considered as small, according to guidelines, it was greater than for the SF-36 component scores (physical: 0.1; mental: 0.1). Sensitivity to change in quality of life measures was greater for OHRQoL than for HRQoL, limiting the usefulness of HRQoL as an outcome measure in dentistry.

Usually, dental treatments work. Not only do clinical parameters of oral disease change, but patients also perceive improvement in their oral health through dental interventions [1, 2]. Because a dental treatment-induced change in oral health can affect so many different aspects of a patient's life, instruments measuring a patient's quality of life (QoL) are often used to quantify the magnitude of treatment effects [3, 4].

Although QoL measures appeal conceptually as the appropriate tools to assess multifaceted change in perceived oral health, practical challenges remain because three broad assessment options exist: (i) QoL, (ii) health-related QoL (HRQoL), and (iii) oral health-related QoL (OHRQoL) instruments. Whereas it is plausible that an individual dental treatment can affect a patient's overall perception of life (=QoL), most dental treatments are not ‘life-changing’. Health-related QoL instruments seem more promising because dental factors and dental interventions have been proven to affect perceived general health [5-7]. Perceived oral health measured together with other health components (cardiovascular health, sleep-related health, etc.) is advantageous in theory, but poses challenges in practice – the change in oral health might not be large enough to be detected using instruments that need to cover a wide range of health conditions with a relatively small number of questions. Oral health-related QoL instruments specifically address this challenge by focusing on oral health concerns; however, the comparability with general health conditions is lost [4, 8]. Consequently, the use of HRQoL measures seems to be most promising as it would allow comparison between dental conditions and interventions with general health conditions and interventions and therefore help to establish oral health as an integral part of general health.

A prerequisite of an instrument when using it to assess the effects of an intervention is the existence of evaluative properties [9], referred to as sensitivity to change. The instrument's score should detect a difference when the underlying construct is expected to change [10]. Whereas sensitivity to change is often used interchangeably with responsiveness, the latter requires not only a statistically significant, but also a clinically meaningful and relevant, change [11]. Others distinguish between internal and external responsiveness [12]. According to this definition, internal responsiveness is equivalent to sensitivity to change.

There are several studies of the sensitivity to change of HRQoL and OHRQoL measures. When using the Short Form-36 (SF-36) [13] for the assessment of changes in HRQoL, sensitivity to change has been demonstrated in different populations of patients with specific general health conditions, such as patients with chronic pain [14] or patients with systemic sclerosis [15]. The same applies to OHRQoL assessments in dentistry, using the Oral Health Impact Profile (OHIP) [16] in specific interventions such as third-molar removal [17], prosthodontic treatments [18, 19], or tooth whitening [20].

However, whether HRQoL measures have sufficient evaluative properties to assess changes caused by dental interventions is less clear. Whilst some authors showed that HRQoL instruments were suitable [21, 22], others reported that HRQoL did not detect changes that were expected to exist [23, 24]. The conditions studied may explain these differences. For chronic pain conditions such as temporomandibular disorder (TMD) [25, 26], and life-threatening conditions such as oral cancer [27, 28], both types of instruments seem applicable, but for typical tooth-related interventions, the situation is less clear because findings from patient populations that represent broad target populations are rare. The findings from edentulous patients provide the first insights [21, 22, 24] but these need to be complemented by studies of patients who receive widely applied dental interventions. Therefore, a head-to-head comparison of HRQoL and OHRQoL measures in a large sample of dental patients receiving typical dental interventions would be informative to assess the relative merits of the two QoL assessment options.

The aim of our study was to investigate and compare sensitivity to change in HRQoL and OHRQoL in patients treated with fixed and removable dentures, and to explore, in that patient population, whether HRQoL measures are sensitive enough to detect changes occurring in perceived oral health, verified by changes in OHRQoL measures.

Material and methods

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. Conflicts of interest
  8. References

A total of 166 consecutive patients, 20–84 yr of age, participated in this prospective intervention study between October 2007 and March 2009 at the Department of Prosthodontics and Materials Science, University of Leipzig. All patients with a need or demand for prosthodontic treatment were included. Patients who were under 18 yr of age or did not understand the questionnaire were excluded from the study. The study protocol (Reg.-No.: 070-2005) was reviewed and approved by the Institutional Review Board of the Medical School, University of Leipzig, Germany. All study participants gave their signed informed consent. Demographic and clinical characteristics at baseline comprised age and gender of the participants, number of teeth and denture status, and global assessments of perceived oral health status and general health status.

Dental undergraduate students provided the prosthodontic treatment, assisted and supervised by experienced dentists. Treatments ranged from single partial crowns over different kinds of fixed partial dentures (FPDs), or removable partial dentures (RPDs), to complete dentures (CDs). However, no implant-supported dentures were provided. The type of prosthodontic device (FPD, RPD, or CD) was recorded at the end of the treatment. All patients completed the self-administered questionnaires on two occasions – before prosthodontic treatment (pretest) and 1 month after treatment was finished (post-test).

Perceived general health was measured using the concept of HRQoL with the German version of the SF-36 [13, 29]. The instrument's 36 items cover eight domains (subscales) of HRQoL, including: physical functioning, role limitations due to physical problems, social functioning, bodily pain, general mental health, role limitations due to emotional problems, vitality, and general health perceptions. These eight domains can form two component (composite) scores: physical (PCS) and mental (MCS) health. The values of the eight domains and of the two component scores were standardized – a value of 50 represents the mean with an SD of 10. Higher values indicate better HRQoL.

Oral health-related quality of life was assessed using the German version of the Oral Health Impact Profile (OHIP-G) [16, 30]. The OHIP-G has 49 items derived from the English-language OHIP and four items specific for the German population: avoid eating with others, take longer to complete meal, joint noises, and dry mouth. Each OHIP question elicits information about how frequently subjects experienced a specific problem in the last month. Responses were made on an ordinal scale: 0 = never, 1 = hardly ever, 2 = occasionally, 3 = often, and 4 = very often. Oral health-related quality of life impairment was characterized by the OHIP-G49 summary score – the sum of the ordinal responses of all 49 items contained in the English-language OHIP (the four German-specific items were omitted to maintain international comparability). For the purpose of clarity, the OHIP-G49 summary score will simply be referred to as ‘OHIP sum score’. As the instrument is a problem index (0-196 OHIP points), higher scores imply greater impairment of OHRQoL.

Reliability of HRQoL measured using the SF-36 and of OHRQoL measured using the OHIP was assessed at baseline by calculating Cronbach's alpha [31], which is a measure of the instrument's internal consistency. According to guidelines [32], reliability of the SF-36 and the OHIP was considered ‘sufficient’ (SF-36: α = 0.95; OHIP: α = 0.97). The strength of the relationship between both QoL constructs was assessed at baseline by calculating Pearson product-moment correlation coefficients. According to Cohen, a coefficient (r) of 0.1 represents a small correlation, 0.3 a medium correlation, and 0.5 a large correlation [33].

Temporal stability of HRQoL measured using the SF-36 and of OHRQoL measured using the OHIP was assessed between two baseline measures in a convenience sample of 116 prosthodontic patients (69.9% of all subjects) using a time interval of about 2 weeks before treatment started. We calculated an intraclass correlation coefficient (ICC) based on a one-way random-effects anova [Shrout & Fleiss ICC (1, 1)] [34], which is the recommended reliability coefficient for the assessment of QoL data [35]. We expected the lower limit of the 95% CI of the ICC to be higher than 0.75, the magnitude of reliability considered ‘excellent’ according to commonly applied guidelines [36]. We compared pairings of both baselines measured using the method of Bland & Altman [37] to quantify intra-individual differences. These statistics represent the test–retest difference expected for 95% of the individuals in the sample.

The SF-36 and OHIP summary and domain scores at baseline and at follow up were tabulated for all patients and then stratified based on the type of treatment. Owing to the limited number of patients in the CD treatment group, all patients receiving RPD or CD treatment were combined (RPD/CD). Differences in SF-36 component and domain scores and in OHIP sum and domain scores between baseline and the follow-up assessment were calculated and tested for statistical significance using the paired t-test. To compare sensitivity to change in both QoL measures, the magnitude of the differences in SF-36 component and OHIP sum scores was judged with respect to the corresponding minimal important difference (MID SF-36: 2–4 points of the standardized scores; MID OHIP-49: 6 points of the summary score) [38, 39]. The effect size for the differences in SF-36 component and OHIP sum scores was calculated, reflecting the impact of the measured difference in relation to the SD [33]. According to Cohen [33], an effect size (Cohen's d) of 0.2 is small, 0.5 is medium, and 0.8 is large. We compared the magnitude of the effect size with values defining a minimal clinically relevant difference in the health measure score (0.5) that was previously determined in a literature review [40].

All analyses were performed using the statistical software package stata (Stata Statistical Software: Release 12. 2011; StataCorp, College Station, TX, USA). The probability of a type I error was set at 0.05.

Data from both questionnaires were complete for 123 (74.1%) patients at baseline and for 125 (75.3%) patients at follow up. The amount of missing information was small (0.3% of the SF-36 and 0.8% of the OHIP-49). Using a regression method, missing values from the OHIP were calculated [41]. Study participants with questionnaires in which more than 10% of the data were missing were excluded from the analyses (one patient; 0.6%), resulting in a total of 165 participants with sufficient data for analyses.

Results

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. Conflicts of interest
  8. References

The mean (SD) age of the study participants was 55.2 (15.8) yr, and about half (51.5%) of the participants were women (Table 1). On average, the participants had 21 teeth. At baseline, 80 (48.5%) participants had no dentures or had FPDs, 78 (47.3%) participants had RPDs, and seven (4.2%) participants had CDs. Most of the participants perceived their oral health and general health as good (51.3% and 51.9%, respectively). During prosthodontic treatment, participants were provided with FPDs (49.1%, = 81) or RPD/CD (RPD: 54.5%, = 75; CD: 5.5%, = 9). Although there was no statistically significant difference in the proportion of women in the FPD and the RPD/CD treatment groups (= 0.308), participants with FPDs were younger, had more teeth, had fewer removable dentures, and rated both their oral and their general health as better than did participants with RPD/CD (all < 0.05).

Table 1. Characteristics of all study participants at baseline and stratified by type of treatment
 All (= 165)Type of treatment
FPD (= 81)RPD/CD (= 84) P
  1. Values are given as mean ± SD or per cent.

  2. FPD, fixed partial denture; RPD/CD, removable partial denture/complete denture.

Demography
Female gender51.555.647.60.308
Age, in years55.2 ± 15.849.6 ± 17.160.7 ± 12.4<0.001
Oral health
Number of teeth20.6 ± 8.625.5 ± 3.614.1 ± 8.9<0.001
Dentures
No/fixed dentures only48.584.014.3<0.001
Removable partial dentures47.316.177.4
Complete dentures4.20.08.3
Self-report
Excellent0.61.30.00.005
Very good6.46.56.3
Good51.361.041.8
Moderate30.127.332.9
Poor11.53.919.0
General health
Self-report
Excellent3.25.31.20.004
Very good18.426.311.0
Good51.948.754.9
Moderate23.417.129.3
Poor3.22.63.7

At baseline, the PCS of the SF-36 for all participants was 49.1 points and the MSC was 51.7 points, indicating a slightly lower perceived physical health status and a somewhat higher perceived mental health status than the reference population used for the standardization of the SF-36 (Table 2). The lowest HRQoL was reported in the general health perception domain (47.8 points) and the highest HRQoL was reported in the vitality domain (52.0 points). This pattern of HRQoL scores was similar across treatment groups. The mean OHIP summary score was 32.9 OHIP points. The OHIP domain scores ranged from 1.7 OHIP points for social disability to 8.1 OHIP points for functional limitations (Table 2). The OHIP summary scores were negatively correlated to the PCS (= −0.32, < 0.001) and the MCS (= −0.37, < 0.001), indicating a medium-strength relationship between OHRQoL and HRQoL. When treatment groups were evaluated, PCS was significantly worse in participants with RPD/CD treatment compared with participants receiving FPDs (47.4 points vs. 51.0 points; independent two-sample t-test: = 0.018). The MCS differences between treatment subgroups were not large and not statistically significant (51.6 points vs. 51.9 points; independent two-sample t-test: = 0.806). In contrast, OHRQoL differed between treatment groups (FPD 42.2 OHIP points vs. RPD/CD 23.3 OHIP points; independent two-sample t-test: < 0.001).

Table 2. Mean scores at baseline and follow-up including SD, and differences in mean scores including 95% CI, of Short Form-36 (SF-36) component and domain scores and Oral Health Impact Profile (OHIP) summary and domain scores with corresponding effect sizes (ES) for all study participants and stratified by type of treatment
 All (= 165)Type of treatment
FPD (= 81)RPD/CD (= 84)
BaselineFollow upDifferenceESBaselineFollow upDifferenceESBaselineFollow upDifferenceES
  1. Values are given as mean ± SD, mean (95% CI) or Cohen's d (effect size).

  2. Values in bold indicate statistically significant changes.

  3. ES, effect size; FPD, fixed partial denture; RPD/CD, removable partial denture/complete denture.

  4. *< 0.05; **< 0.01; ***< 0.001.

SF-36
Physical component summary score49.1 ± 9.650.2 ± 9.4 1.0 (0.1 to 2.0) * 0.1 51.0 ± 9.252.2 ± 9.71.3 (−0.1 to 2.6)0.147.4 ± 9.848.2 ± 8.70.8 (−0.5 to 2.2)0.1
Mental component summary score51.7 ± 8.651.2 ± 10.0−0.5 (−1.7 to 0.6)0.151.9 ± 8.850.1 ± 10.5−1.7 (−3.5 to 0.0)0.251.6 ± 8.352.2 ± 9.50.6 (−1.0 to 2.2)0.1
Physical functioning50.0 ± 8.950.3 ± 9.40.3 (−0.6 to 1.2)0.051.9 ± 8.251.9 ± 8.90.0 (−1.2 to 1.1)0.048.2 ± 9.348.8 ± 9.70.6 (−0.8 to 2.1)0.1
Role limitations due to physical problems49.2 ± 7.949.7 ± 7.60.5 (−0.6 to 1.6)0.150.3 ± 8.151.1 ± 7.40.8 (−0.9 to 2.6)0.148.1 ± 7.648.3 ± 7.50.2 (−1.3 to 1.6)0.0
Bodily pain50.8 ± 11.352.4 ± 10.0 1.6 (0.2 to 2.9) * 0.1 51.7 ± 10.853.2 ± 9.81.5 (−0.3 to 3.3)0.150.0 ± 11.851.6 ± 10.21.7 (−0.4 to 3.7)0.2
General health perception47.8 ± 10.548.4 ± 10.20.6 (−0.5 to 1.7)0.150.0 ± 10.350.4 ± 10.20.3 (−1.4 to 2.1)0.045.6 ± 10.346.6 ± 9.90.9 (−0.4 to 2.2)0.1
Vitality52.0 ± 9.451.7 ± 9.3−0.3 (−1.6 to 0.9)0.053.1 ± 9.251.6 ± 9.4−1.6 (−3.5 to 0.4)0.251.0 ± 9.651.8 ± 9.30.8 (−0.6 to 2.3)0.1
Social functioning51.7 ± 9.051.9 ± 9.30.1 (−1.3 to 1.6)0.052.0 ± 9.251.6 ± 9.6−0.3 (−2.2 to 1.5)0.051.5 ± 8.952.1 ± 9.10.6 (−1.6 to 2.8)0.1
Role limitations due to emotional problems51.1 ± 7.050.9 ± 7.3−0.3 (−1.2 to 0.7)0.051.9 ± 7.251.2 ± 7.6−0.7 (−2.1 to 0.8)0.150.4 ± 6.850.5 ± 7.10.1 (−1.1 to 1.3)0.0
General mental health50.5 ± 9.250.3 ± 10.9−0.2 (−1.6 to 1.2)0.050.9 ± 9.049.3 ± 11.3−1.6 (−3.6 to 0.5)0.250.1 ± 9.351.2 ± 10.51.1 (−0.7 to 2.9)0.1
OHIP-49
Summary score32.9 ± 28.526.2 ± 25.3 −6.7 (−9.9 to −3.4) *** 0.2 23.3 ± 21.519.3 ± 18.8 −4.0 (−7.8 to −0.2) * 0.2 42.2 ± 31.333.0 ± 28.8 −9.2 (−14.5 to −3.9) *** 0.3
Functional limitations8.1 ± 5.95.9 ± 5.1 −2.2 (−2.9 to −1.5) *** 0.4 5.8 ± 4.54.3 ± 3.7 −1.5 (−2.3 to −0.6) *** 0.4 10.3 ± 6.27.4 ± 5.8 −2.9 (−4.0 to −1.8) *** 0.5
Physical pain7.1 ± 6.16.8 ± 5.8−0.2 (−1.1 to 0.6)0.05.6 ± 5.55.0 ± 4.4−0.5 (−1.7 to 0.6)0.18.5 ± 6.48.6 ± 6.50.0 (−1.3 to 1.3)0.0
Psychological discomfort4.2 ± 4.32.8 ± 3.3 −1.4 (−2.0 to −0.8) *** 0.4 3.3 ± 3.52.4 ± 2.6 −0.8 (−1.6 to −0.1) * 0.3 5.2 ± 4.83.2 ± 3.9 −2.0 (−2.9 to −1.0) *** 0.5
Physical disability5.2 ± 6.34.1 ± 5.3 −1.1 (−1.9 to −0.4) ** 0.2 2.6 ± 3.62.3 ± 3.3−0.2 (−1.0 to 0.5)0.17.8 ± 7.25.8 ± 6.2 −2.0 (−3.3 to −0.7) ** 0.3
Psychological disability3.6 ± 4.12.8 ± 3.7 −0.8 (−1.3 to −0.3) ** 0.2 2.6 ± 3.52.2 ± 3.0−0.4 (−1.1 to 0.2)0.14.5 ± 4.43.4 ± 4.2 −1.1 (−1.9 to −0.4) ** 0.3
Social disability1.7 ± 2.71.4 ± 2.3−0.3 (−0.7 to 0.0)0.11.0 ± 1.91.0 ± 1.9−0.1 (−0.5 to 0.3)0.02.3 ± 3.11.8 ± 2.7 −0.6 (−1.1 to 0.0) * 0.2
Handicap3.1 ± 3.62.5 ± 3.3 −0.6 (−1.0 to −0.1) * 0.2 2.5 ± 3.22.1 ± 3.0−0.4 (−1.1 to 0.2)0.13.6 ± 3.92.9 ± 3.6−0.7 (−1.4 to 0.0)0.2

At baseline, temporal stability (test–retest reliability) of HRQoL was excellent for the PCS (ICC = 0.85; 95% CI: 0.80–0.90) and fair to good for the MCS (ICC = 0.66; 95% CI: 0.55–0.76), but individual differences between both baseline measures (indicated by limits of agreement) were of considerable magnitude (PCS: −10.8 to 11.5; MCS: −13.3 to 13.7). The temporal stability of OHRQoL (OHIP sum score) was excellent (ICC = 0.86; 95% CI: 0.82–0.91), but limits of agreement were also wide (−31.9 to 32.7 OHIP points). The mean differences between both baseline measures of the PCS (0.3 points), the MCS (0.2 points), and the OHIP sum score (0.4 points) were not statistically significant (paired t-test: all > 0.05).

After prosthodontic treatment, the PCS was 50.2 points and the MCS was 51.2 points (Table 2), very close to the value of 50 points in the reference population. The mean OHIP summary score at follow up was 26.2 points, indicating still a substantial impairment in OHRQoL. Similar to the findings at baseline, patients with RPD/CD treatment had at follow-up worse PCS than did patients with FPD therapy. Oral health-related QoL was still more impaired after RPD/CD treatment than after FPD therapy (33.0 OHIP points vs. 19.3 OHIP points; independent two sample t-test: < 0.001).

Health-related QoL improved during prosthodontic treatment, indicated by a slight, but statistically significant, increase in PCS at follow up (difference: 1.0 points; paired t-test: = 0.034; Table 2). Although the direction of differences was identical in both treatment groups, these differences were not statistically significant (paired t-test: both > 0.05). The increase in PCS was mainly caused by a score increase in the bodily pain domain (1.6; paired t-test: = 0.023), corresponding to a decrease in perceived pain. The MCS did not change substantially during treatment (paired t-test; = 0.355). Improvement in OHRQoL during prosthodontic treatments (difference in OHIP sum score: −6.7 points) was statistically significant (paired t-test: < 0.001). The change in OHIP sum scores was smaller in magnitude in FPD patients than in RPD/CD patients, but both differences were statistically significant (paired t-test: both < 0.05). The scores of almost all OHRQoL domains decreased substantially following prosthodontic treatment. Only the scores in physical pain and social disability stayed virtually identical to baseline scores.

When using effect size for the interpretation of QoL changes, PCS and MCS did not reach a ‘small’ effect (both effects = 0.1), but OHIP scores did (0.2). When using MID for the interpretation of QoL changes, PCS and MCS did not show a clinically relevant change compared with an MID of 2–4 points, but OHIP scores showed a clinically relevant change for all patients, including the RPD/CD treatment group, when compared with an MID of 6 points.

Discussion

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. Conflicts of interest
  8. References

This study compared sensitivity to change in HRQoL and OHRQoL measures in patients treated with fixed, removable, and complete dentures and evaluated whether HRQoL measures were sensitive enough to detect changes in perceived oral health, verified by changes in OHRQoL measures. Sensitivity to change of QoL measures was higher in OHRQoL, assessed using the OHIP, than in HRQoL, assessed using the SF-36. Whereas OHRQoL improved substantially and was clinically relevant by exceeding the threshold of the MID of the OHIP (6 points) [39], HRQoL changed only slightly and did not exceed the MID of the SF-36 (2–4 points) [38]. In other words, on average, patients perceived the effect of treatments as an improvement of their oral health but not as an improvement of their general health, including oral health as a component.

Owing to the different calculation methods for both instruments (summary scores for the OHIP and standardized scores for the SF-36), comparability of the magnitude of score changes is limited. However, effect sizes, unitless values relating magnitude of change scores to the variability of change of the measures, are available and were higher for the OHIP sum scores than for the SF-36 component scores. Furthermore, changes in OHIP sum scores were statistically significant in all subgroups, whereas for the SF-36, neither the change in physical component scores nor the change in mental component scores were significant in the patient subgroups. This situation, that treatment effects known to exist are reflected in the OHRQoL measure but not consistently in the HRQoL measure, limits the utility of the HRQoL as an outcome measure in dentistry.

Although an improvement in HRQoL unrelated to chance was observed, this change was substantially smaller in magnitude than the improvement in OHRQoL. Pains in the orofacial area, aspects of tooth decay that cause symptoms, and tooth loss are only some of the clinical conditions that impact patients often in a broader way, and, furthermore, impair not only oral health, but also general health and daily performance [5, 42-45]. Therefore, treatment of these conditions should result in improved perceived general health and well-being and consequently the finding of just a small change in HRQoL during prosthodontic treatment could be expected. On the other hand, one would also expect that HRQoL and OHRQoL are compatible, but when comparing changes in the domains of HRQoL and OHRQoL, the results were not consistent. For HRQoL, there was a significant change (improvement) in the bodily pain domain. In contrast, although the OHIP summary score decreased substantially, indicating improved OHRQoL, the OHIP domain score of physical pain stayed virtually the same. This was surprising given the finding of significant HRQoL improvements in perceived bodily pain. Pain in the orofacial area should be part of perceived bodily pain.

The study's findings can be compared with those of other studies investigating the effect of comprehensive oral rehabilitations on HRQoL or OHRQoL. Although our findings are in line with those of other studies using OHRQoL as an outcome measure, which show that dental treatments improved OHRQoL [22, 24, 46, 47], findings of studies using HRQoL measures are conflicting. Prosthodontic treatment increased a global rating of perceived general health [48]. Heydecke et al. [22] found that in patients who were provided with conventional complete dentures, HRQoL did not improve substantially. Only patients provided with implant-supported complete dentures perceived improvements in social functioning, role limitations due to emotional problems, and vitality. Other authors observed improvements in the general health dimension of the SF-36 in patients provided with conventional complete dentures [24]. In another study comparing conventional and implant dentures, limitations as a result of emotional problems and social functioning, both corresponding to the mental health component of the SF-36, improved in patients who requested and received conventional dentures, but not in patients who received implant-retained dentures [21]. Finally, a meta-analysis of randomized-controlled trials summarized the impact of implant support for mandibular dentures and concluded that there is a lack of evidence for an effect on HRQoL [49]. Some findings were also available for other dental interventions. For example, orthodontic treatment did not result in an increase in well-being [23]. Overall, these contradictory findings of the studies for the effect of dental treatments on HRQoL indicate that the size of the effect is still uncertain, but is probably very small and therefore difficult to be consistently detected with the HRQoL measures available. A lesser ability to detect differences in oral health using the SF-36 compared with the OHIP has also been shown in cross-sectional studies [50, 51]. However, HRQoL seems to perform as well as OHRQoL in conditions with more impact on general health, including chronic pain conditions, such as TMD [25, 26], and life-threatening conditions, such as oral cancer [27, 28].

This study has strengths and limitations. A validated German version [29] of the SF-36 [13], a well-accepted and widely distributed measure for HRQoL, was applied. Furthermore, the validated German version [30] of the OHIP [16] is a measure with known sensitivity to change in perceived oral health [46]. Whether the study's findings are generalizable to other HRQoL and OHRQoL measures is not known, but many of these instruments are similar.

In our study, typical prosthodontic patients with a wide range of prosthodontic treatments, ranging from FDPs to CDs, were included. However, no implants were placed in the study patients and no implant-supported dentures were incorporated. Therefore, the results may not be generalized to dental implant patients. The socio-economic status of patients was not assessed. Even though socio-economic characteristics are associated with subjects' health perceptions [8], they should not affect change scores.

Treatment effects were assessed 4 weeks after the completion of prosthodontic treatment. Therefore, we captured short-term effects. Assuming that improved oral health affects HRQoL by changes in general health and in health behaviour (such as nutrition), it might be some time before the patients perceive improvements in their HRQoL. Studies with longer follow-up periods might be necessary to assess delayed, long-term improvements in HRQoL after prosthodontic treatments.

Methodological factors might influence the magnitude of (O)HRQoL changes. Sometimes, the actual true change of (O)HRQoL might be confounded with the way in which it is assessed because patients may change their internal standards of QoL during treatments. This phenomenon is called response shift [52]. It has been found that HRQoL [53, 54] and OHRQoL [55, 56] measures can be affected by response shift. We found response shift in a previous investigation in the same patient population when OHRQoL change was assessed retrospectively [56]. However, it is not clear whether response shift is solely a result of changes in patients' internal standards, values, and concepts of QoL during treatments, or if it results from psychological processes such as cognitive dissonance or implicit theory of change in retrospective baseline assessments [57, 58] and therefore is some form of a placebo effect [59]. Even though treatment effects might be slightly higher in retrospective studies than they are in prospective studies, these prospective studies are currently the standard in clinical research. Consequently, it seems justified to determine sensitivity to change in studies with a prospective assessment of treatment effects and not for a retrospective assessment with consideration of potential response shift effects.

Regression to the mean is another factor that can influence the magnitude of (O)HRQoL change. This phenomenon appears when study participants are selected on the basis of how low or how high the values of their outcome variables (such as (O)HRQoL) are. In such situations, participants change to higher or lower values of the outcome variable just because of measurement error. Accordingly, they gravitate to the middle of the data distribution at follow up, purely for statistical reasons. However, participants were recruited based on clinical considerations and treatment demand; a high outcome measure at baseline was not an inclusion criterion. Furthermore, both HRQoL and OHRQoL were quite stable before treatment started with no significant shift to a specific direction. Therefore, it can be assumed that changes in HRQoL and OHRQoL were not affected by regression to the mean.

Our study design is an intervention study using a consecutive series of patients, which, in general, does not lead to strong causal inference. However, determination of sensitivity of change does not rely on causal inference because it describes the magnitude of change in relation to the variability of the change, or a criterion such as the MID. Many conditions affecting oral and general health and, consequently, HRQoL and OHRQoL, cannot be controlled in such a study. For ethical reasons, no control group of patients without prosthodontic treatment could be included. Normal fluctuation of HRQoL and OHRQoL above and beyond statistical variation may explain our findings because (O)HRQoL is a dynamic construct [60].

Based on clinical expertise, prosthodontic treatment is efficacious with an effect that is not usually small. Therefore, instruments measuring this effect should have sensitivity to change and responsiveness. However, treatment effects on HRQoL, 4 weeks after prosthodontic treatment was finished, were small and limit the usefulness of HRQoL as an outcome measure in dentistry. Oral health-related QoL measures seem to provide more easily interpretable results and should be preferred when assessing treatment effects in typical dental patients.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. Conflicts of interest
  8. References

The authors are grateful to Ms Annett Schrock (University of Leipzig) for help with data management and to Ms Andrea Medina (University of Minnesota) for valuable comments on an earlier version of the manuscript.

References

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. Conflicts of interest
  8. References
  • 1
    John MT, Reissmann DR, Allen F, Biffar R. The short-term effect of prosthodontic treatment on self-reported oral health status: the use of a single-item questionnaire. Int J Prosthodont 2007; 20: 507513.
  • 2
    Gibson G, Reifenstahl EF, Wehler CJ, Rich SE, Kressin NR, King TB, Jones JA. Dental treatment improves self-rated oral health in homeless veterans–a brief communication. J Public Health Dent 2008; 68: 111115.
  • 3
    Hultin M, Davidson T, Gynther G, Helgesson G, Jemt T, Lekholm U, Nilner K, Nordenram G, Norlund A, Rohlin M, Sunnegardh-Gronberg K, Tranaeus S. Oral rehabilitation of tooth loss: a systematic review of quantitative studies of OHRQoL. Int J Prosthodont 2012; 25: 543552.
  • 4
    Allen PF. Assessment of oral health related quality of life. Health Qual Life Outcomes 2003; 1: 40.
  • 5
    Naito M, Yuasa H, Nomura Y, Nakayama T, Hamajima N, Hanada N. Oral health status and health-related quality of life: a systematic review. J Oral Sci 2006; 48: 17.
  • 6
    Richmond S, Chestnutt I, Shennan J, Brown R. The relationship of medical and dental factors to perceived general and dental health. Community Dent Oral Epidemiol 2007; 35: 8997.
  • 7
    Brennan DS, Spencer AJ, Roberts-Thomson KF. Tooth loss, chewing ability and quality of life. Qual Life Res 2008; 17: 227235.
  • 8
    Sischo L, Broder HL. Oral health-related quality of life: what, why, how, and future implications. J Dent Res 2011; 90: 12641270.
  • 9
    Guyatt GH, Kirshner B, Jaeschke R. Measuring health status: what are the necessary measurement properties? J Clin Epidemiol 1992; 45: 13411345.
  • 10
    Liang MH. Longitudinal construct validity: establishment of clinical meaning in patient evaluative instruments. Med Care 2000; 38: II84II90.
  • 11
    Epstein RS. Responsiveness in quality-of-life assessment: nomenclature, determinants, and clinical applications. Med Care 2000; 38: II91II94.
  • 12
    Husted JA, Cook RJ, Farewell VT, Gladman DD. Methods for assessing responsiveness: a critical review and recommendations. J Clin Epidemiol 2000; 53: 459468.
  • 13
    Ware JE Jr, Sherbourne CD. The MOS 36-item short-form health survey (SF-36). I. Conceptual framework and item selection. Med Care 1992; 30: 473483.
  • 14
    Wittink H, Turk DC, Carr DB, Sukiennik A, Rogers W. Comparison of the redundancy, reliability, and responsiveness to change among SF-36, Oswestry Disability Index, and Multidimensional Pain Inventory. Clin J Pain 2004; 20: 133142.
  • 15
    Khanna D, Furst DE, Clements PJ, Park GS, Hays RD, Yoon J, Korn JH, Merkel PA, Rothfield N, Wigley FM, Moreland LW, Silver R, Steen VD, Weisman M, Mayes MD, Collier DH, Medsger TA Jr, Seibold JR. Responsiveness of the SF-36 and the Health Assessment Questionnaire Disability Index in a systemic sclerosis clinical trial. J Rheumatol 2005; 32: 832840.
  • 16
    Slade GD, Spencer AJ. Development and evaluation of the Oral Health Impact Profile. Community Dent Health 1994; 11: 311.
  • 17
    Kieffer JM, van Wijk AJ, Ho JP, Lindeboom JA. The internal responsiveness of the Oral Health Impact Profile-14 to detect differences in clinical parameters related to surgical third molar removal. Qual Life Res 2012; 21: 12411247.
  • 18
    Allen PF, McMillan AS, Locker D. An assessment of sensitivity to change of the Oral Health Impact Profile in a clinical trial. Community Dent Oral Epidemiol 2001; 29: 175182.
  • 19
    Awad M, Al-Shamrany M, Locker D, Allen F, Feine J. Effect of reducing the number of items of the Oral Health Impact Profile on responsiveness, validity and reliability in edentulous populations. Community Dent Oral Epidemiol 2008; 36: 1220.
  • 20
    McGrath C, Wong AH, Lo EC, Cheung CS. The sensitivity and responsiveness of an oral health related quality of life measure to tooth whitening. J Dent 2005; 33: 697702.
  • 21
    Allen PF, McMillan AS. A longitudinal study of quality of life outcomes in older adults requesting implant prostheses and complete removable dentures. Clin Oral Implants Res 2003; 14: 173179.
  • 22
    Heydecke G, Locker D, Awad MA, Lund JP, Feine JS. Oral and general health-related quality of life with conventional and implant dentures. Community Dent Oral Epidemiol 2003; 31: 161168.
  • 23
    Arrow P, Brennan D, Spencer AJ. Quality of life and psychosocial outcomes after fixed orthodontic treatment: a 17-year observational cohort study. Community Dent Oral Epidemiol 2011; 39: 505514.
  • 24
    Kuo HC, Kuo YS, Lee IC, Wang JC, Yang YH. The association of responsiveness in oral and general health-related quality of life with patients' satisfaction of new complete dentures. Qual Life Res 2013; 22: 16651674.
  • 25
    Bernhardt O, Gesch D, Schwahn C, Mack F, Meyer G, John U, Kocher T. Risk factors for headache, including TMD signs and symptoms, and their impact on quality of life. Results of the Study of Health in Pomerania (SHIP). Quintessence Int 2005; 36: 5564.
  • 26
    Reissmann DR, John MT, Schierz O, Wassell RW. Functional and psychosocial impact related to specific temporomandibular disorder diagnoses. J Dent 2007; 35: 643650.
  • 27
    Chandu A, Smith AC, Rogers SN. Health-related quality of life in oral cancer: a review. J Oral Maxillofac Surg 2006; 64: 495502.
  • 28
    Hassel AJ, Danner D, Freier K, Hofele C, Becker-Bikowski K, Engel M. Oral health-related quality of life and depression/anxiety in long-term recurrence-free patients after treatment for advanced oral squamous cell cancer. J Craniomaxillofac Surg 2012; 40: e99e102.
  • 29
    Bullinger M. German translation and psychometric testing of the SF-36 Health Survey: preliminary results from the IQOLA Project. International Quality of Life Assessment. Soc Sci Med 1995; 41: 13591366.
  • 30
    John MT, Patrick DL, Slade GD. The German version of the Oral Health Impact Profile–translation and psychometric properties. Eur J Oral Sci 2002; 110: 425433.
  • 31
    Cronbach LJ. Coefficient alpha and the internal structure of tests. Psychometrika 1951; 16: 297334.
  • 32
    Bland JM, Altman DG. Cronbach's alpha. BMJ 1997; 314: 572.
  • 33
    Cohen J. Statistical power analysis for the behavioral sciences, 2nd edn. Hillsdale, NJ: Lawrence Earlbaum Associates, 1988.
  • 34
    Shrout PE, Fleiss JL. Intraclass correlations: uses in assessing rater reliability. Psychol Bull 1979; 86: 420428.
  • 35
    Schuck P. Assessing reproducibility for interval data in health-related quality of life questionnaires: which coefficient should be used? Qual Life Res 2004; 13: 571586.
  • 36
    Fleiss JL, Levin B, Paik MC. Statistical methods for rates and proportions. Hoboken, NJ: John Wiley & Sons, 2003.
  • 37
    Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986; 1: 307310.
  • 38
    Swigris JJ, Brown KK, Behr J, du Bois RM, King TE, Raghu G, Wamboldt FS. The SF-36 and SGRQ: validity and first look at minimum important differences in IPF. Respir Med 2010; 104: 296304.
  • 39
    John MT, Reissmann DR, Szentpetery A, Steele J. An approach to define clinical significance in prosthodontics. J Prosthodont 2009; 18: 455460.
  • 40
    Jaeschke R, Singer J, Guyatt GH. Measurement of health status. Ascertaining the minimal clinically important difference. Control Clin Trials 1989; 10: 407415.
  • 41
    John MT, LeResche L, Koepsell TD, Hujoel P, Miglioretti DL, Micheelis W. Oral health-related quality of life in Germany. Eur J Oral Sci 2003; 111: 483491.
  • 42
    Mack F, Schwahn C, Feine JS, Mundt T, Bernhardt O, John U, Kocher PT, Biffar R. The impact of tooth loss on general health related to quality of life among elderly Pomeranians: results from the study of health in Pomerania (SHIP-O). Int J Prosthodont 2005; 18: 414419.
  • 43
    Shimada A, Sasaki Y, Mataki S. An evaluation of self-reported oral health and health-related quality of life. J Med Dent Sci 2005; 52: 6572.
  • 44
    Akifusa S, Soh I, Ansai T, Hamasaki T, Takata Y, Yohida A, Fukuhara M, Sonoki K, Takehara T. Relationship of number of remaining teeth to health-related quality of life in community-dwelling elderly. Gerodontology 2005; 22: 9197.
  • 45
    Adulyanon S, Vourapukjaru J, Sheiham A. Oral impacts affecting daily performance in a low dental disease Thai population. Community Dent Oral Epidemiol 1996; 24: 385389.
  • 46
    John MT, Slade GD, Szentpetery A, Setz JM. Oral health-related quality of life in patients treated with fixed, removable, and complete dentures 1 month and 6 to 12 months after treatment. Int J Prosthodont 2004; 17: 503511.
  • 47
    Wolfart S, Moll D, Hilgers RD, Wolfart M, Kern M. Implant placement under existing removable dental prostheses and its effect on oral health-related quality of life. Clin Oral Implants Res 2013; 24: 13541359.
  • 48
    Reissmann DR, Schierz O, Szentpetery AG, John MT. Improved perceived general health is observed with prosthodontic treatment. J Dent 2011; 39: 326331.
  • 49
    Emami E, Heydecke G, Rompre PH, de Grandmont P, Feine JS. Impact of implant support for mandibular dentures on satisfaction, oral and general health-related quality of life: a meta-analysis of randomized-controlled trials. Clin Oral Implants Res 2009; 20: 533544.
  • 50
    Allen PF, McMillan AS, Walshaw D, Locker D. A comparison of the validity of generic- and disease-specific measures in the assessment of oral health-related quality of life. Community Dent Oral Epidemiol 1999; 27: 344352.
  • 51
    Broder HL, Slade G, Caine R, Reisine S. Perceived impact of oral health conditions among minority adolescents. J Public Health Dent 2000; 60: 189192.
  • 52
    Sprangers MA, Schwartz CE. Integrating response shift into health-related quality of life research: a theoretical model. Soc Sci Med 1999; 48: 15071515.
  • 53
    Razmjou H, Schwartz CE, Yee A, Finkelstein JA. Traditional assessment of health outcome following total knee arthroplasty was confounded by response shift phenomenon. J Clin Epidemiol 2009; 62: 9196.
  • 54
    Schwartz CE, Finkelstein JA. Understanding inconsistencies in patient-reported outcomes after spine treatment: response shift phenomena. Spine J 2009; 9: 10391045.
  • 55
    Ring L, Hofer S, Heuston F, Harris D, O'Boyle CA. Response shift masks the treatment impact on patient reported outcomes (PROs): the example of individual quality of life in edentulous patients. Health Qual Life Outcomes 2005; 3: 55.
  • 56
    Reissmann DR, Remmler A, John MT, Schierz O, Hirsch C. Impact of response shift on the assessment of treatment effects using the Oral Health Impact Profile. Eur J Oral Sci 2012; 120: 520525.
  • 57
    Norman G. Hi! How are you? Response shift, implicit theories and differing epistemologies. Qual Life Res 2003; 12: 239249.
  • 58
    Festinger L. A theory of social comparison processes. Hum Relat 1954; 7: 117140.
  • 59
    Wilson IB. Clinical understanding and clinical implications of response shift. Soc Sci Med 1999; 48: 15771588.
  • 60
    Allison PJ, Locker D, Feine JS. Quality of life: a dynamic construct. Soc Sci Med 1997; 45: 221230.