Abstract
Objective: To test validity and reliability of the Spanish version of the 5-item Oral Health Impact Profile (OHIP-Sp5). Methods: Spanish-speaking dental patients (n = 331, response rate = 61%, age: 42.9 + 12.3 years, 59% female) with a scheduled appointment at HealthPartners dental clinics in Minnesota, USA, were investigated. To assess score reliability, we computed Cronbach’s alpha, expecting ‘good’ reliability (alpha > 0.70). To assess score validity, we correlated the OHIP-Sp5 summary score with five OHRQoL measures [49- and 14-item OHIP-Sp, the Spanish version of the General Oral Health Assessment Index (GOHAI-Sp), the Spanish version of the Oral Impacts on Daily Performances (OIDP-Sp) and a Global Oral Health Assessment]. We expected a pattern of ‘very large’ (r > 0.70) correlation coefficients for OHIP-Sp5 relationships with the two longer OHIP-Sp versions and ‘large’ (r > 0.50) correlation coefficients for the other three measures. Results: Patients had a mean OHIP-Sp5 score of 3.7 (SD = 4.0). The Instrument’s reliability was, as expected, ‘good’, according to the Cronbach’s alpha statistic of 0.83. The Instrument’s validity was supported by the expected pattern of validity coefficients. OHIP-Sp5 summary scores correlated with OHIP-Sp14 as well as with OHIP-Sp49 (both r = 0.95: ‘very large’ effect) and with GOHAI-Sp, OIDP and the Global Oral Health Assessment with r = −0.80, 0.73, and −0.56 (absolute effect magnitude all ‘very large’ or ‘large’), respectively. Conclusions: The Spanish version of OHIP-5 is a brief and psychometrically sound instrument to measure oral health-related quality of life (OHRQoL) in Spanish-speaking populations. It can effectively replace longer OHIP instruments and would be applicable across all settings of clinical practice and research.
Key words: Validation studies, psychometrics, quality of life, patient outcome assessment
INTRODUCTION
Oral health-related quality of life (OHRQoL) is the most widely used dental patient-reported outcome (dPRO) 1 in evidence-based dentistry2 and value-based oral health care3. dPRO instruments are used by both researchers and oral health practitioners to evaluate the impact of dental interventions on patients’ quality of life4. Over the years, numerous questionnaires have been developed to measure the dPRO, OHRQoL. A recent systematic review of dPRO measures (dPROMs) identified 20 different questionnaires containing 36 unique dPROs. While the questionnaires varied in length, it was determined that oral health-generic dPROMs quantified OHRQoL using essentially four major attributes (i.e., dimensions): Oral Function; Orofacial Appearance; Orofacial Pain; and Psychosocial Impact5. These dimensions, the core components of OHRQoL, are most parsimoniously measured using the five-item Oral Health Impact Profile (OHIP-5), a validated instrument with one item representing at least each of the four dimensions6, 7. This dPROM constitutes a low-burden assessment tool for both patients and health professionals8, and it allows dentists to assess patient-reported outcomes that would be compatible with OHRQoL assessments used in research settings. The OHIP-5 instrument is available in several languages, including German9, Japanese10, Dutch11, Swedish8 and English12. However, a Spanish version of the OHIP-5 has not yet been developed. With more than 20 countries having primarily a Spanish-speaking population and more than 414 million people speaking Spanish worldwide13, this is the second most widely spoken native language, even surpassing English as a native language. In addition, in recent years Spanish has been reported to be one of the most rapidly growing languages, both written and spoken13.
The aim of this study was to develop a Spanish version of the 5-item Oral Health Impact Profile (OHIP-Sp5) and determine the reliability and validity of the composite measurement score in Spanish-speaking dental patients.
METHODS
Participants, study design and setting
In this cross-sectional study, a total of 2,113 English- and Spanish-speaking adult dental patients from HealthPartners dental clinics in Minnesota, USA, were targeted. HealthPartners (www.healthpartners.com) is a large upper-Midwestern integrated health system that consists of a dental group, a medical group (including hospitals), a medical plan and a dental plan. The HealthPartners medical group serves more than one million patients, and the HealthPartners dental group serves over 200,000 patients.
Starting in July 2014, patients 18 years or older who had a scheduled dental appointment were approached to participate in the study. A consecutive sample of patients was targeted at each participating HealthPartners dental clinic. A battery of dPROMs, among them Spanish versions of OHRQoL instruments, such as the Oral Health Impact Profile (OHIP-Sp)14, Oral Impacts on Daily Performances (OIDP-Sp)15, and General Oral Health Assessment Index (GOHAI-Sp)16, 17, were delivered to patients. Patients filled out the self-administered questionnaires at home and sent the questionnaires back to HealthPartners Institute. They received $50 compensation. Recruitment ended in April 2016.
To identify Spanish-speaking dental patients, the language indicator in the patients’ electronic dental record or in the electronic medical record was used. Spanish speaking patients could also be bilingual. These bilingual patients were identified by checking their country of origin in the electronic health record. The patient was considered as bilingual if their language was English or this information was missing from the electronic health record and their country of origin was from the list of countries we identified as Spanish speaking: Argentina, Bolivia, Chile, Colombia, Costa Rica, Cuba, the Dominican Republic, Ecuador, El Salvador, Equatorial Guinea, Guatemala, Honduras, Mexico, Nicaragua, Panama, Paraguay, Peru, Spain, Uruguay and/or Venezuela. Bilingual patients were offered an English version instead of the Spanish version if this was preferred. Accordingly, bilingual patients self-selected English or Spanish. Spanish-speaking patients using Spanish language forms (n = 540) were included in the study, and 331 provided information (response rate = 61.3%, age: 42.9 ± 12.3 years, 59% female). The study was reviewed and approved by the Institutional Review Board of HealthPartners Institute (Study Number A11-136) and was conducted in full accordance with the World Medical Association Declaration of Helsinki. Written informed consent was obtained from participants.
The 5-item OHIP – Spanish Version
The 5-item OHIP was first proposed in Germany9. It was developed using best subset regression. This regression technique searched through all possible combinations of the 49 available questionnaire items to select a pool of items that performs ‘best’ according to a defined statistical criterion (i.e., in the case of OHIP-5, to explain at least 90% OHIP-G [OHIP-German version] summary score variance). The technique evaluated how well all 49 one-item, 1,176 two-item, 18,424 three-item combinations etc., performed according to the previous criterion. Among the 1,906,884 five-OHIP item combinations, the item set that performed best in terms of both statistical properties and coverage of subject matter, as well as interpretability, was chosen as the final version of the 5-item OHIP 9. Consequently, this ultrashort version contains only 10% of the items but captures about 90% of the score information compared with the 49-item version9, 12. As it contains at least one indicator for each of the four OHRQoL dimensions Oral Function, Orofacial Pain, Orofacial Appearance, and Psychosocial Impact6, 7, the instrument has excellent content validity while simultaneously minimising burden.
A Spanish version of the OHIP-49 (OHIP-Sp49) does exist14. We used it as the source of items for the OHIP-Sp5 (Table 1) and used the past month as the recall period.
Table 1.
Items included and domain distribution in the Spanish 5-item version of the Oral Health Impact Profile (OHIP-Sp5)
| English | Spanish | OHRQoL domain | |
|---|---|---|---|
| 1 | Have you had difficulty chewing any foods because of problems with your teeth, mouth, dentures or jaw? | ¿Has tenido dificultades mordiendo algún alimento por problemas con tus dientes, boca o prótesis? | Oral function |
| 2 | Have you had painful aching in your mouth? | ¿Has tenido molestias dolorosas en tu boca? | Orofacial pain |
| 3 | Have you felt uncomfortable about the appearance of your teeth, mouth, dentures or jaw? | ¿Has sentido disconfort sobre la apariencia de tus dientes, boca o prótesis? | Orofacial appearance |
| 4 | Have you felt that there has been less flavour in your food because of problems with your teeth, mouth, dentures or jaws? | ¿Has sentido que hay menos sabor en tus alimentos por problemas con tus dientes, boca o prótesis? | Oral function |
| 5 | Have you had difficulty doing your usual jobs because of problems with your teeth, mouth, dentures or jaws? | ¿Has tenido dificultades haciendo tu trabajo habitual por problemas con tus dientes, boca o prótesis? | Psychosocial impact |
Responses were presented on a 5-point ordinal scale (0, never; 1, hardly ever; 2, occasionally; 3, fairly often; and 4, very often). Summing all responses resulted in a score ranging from a minimum of 0 to a maximum of 20. A larger score indicated more negative impacts of oral health problems.
Assessment of reliability and validity of OHIP-Sp5 in dental patients
Reliability
To assess reliability of the OHIP-Sp5, an internal consistency analysis was performed by computing the Cronbach’s alpha statistic18. Estimate values were interpreted according to Kline as follows: acceptable (0.60–0.69); good (0.70–0.89); excellent (>0.90)19. In addition, item-rest correlations were reported for each OHIP-Sp5 item. In a reliable scale, these values should all be high and similar across items.
Validity
To assess validity of the OHIP-Sp5, we investigated the score’s relationship with other measures of the same construct (i.e., OHRQoL). Most studies in the OHRQoL domain to assess validity of dPROMs have included only one additional measure12, 14. However, with the aim to be comparable with many studies, we used five OHRQoL measures: OHIP-Sp4914; OHIP-Sp1420; GOHAI-Sp16, 17; OIDP-Sp15; and a global assessment of oral health. Table 2 depicts a summary of number of items, response options, and the minimum and maximum values possible for each dPROM/measure.
Table 2.
dPROMs) used to assess validity of the Spanish 5-item version of the Oral Health Impact Profile (OHIP-Sp5)
To characterise how OHIP-Sp5 and the five validity measures were related, scatterplots and LOcally WEighted Scatterplot Smoothing (LOWESS) were used. LOWESS fits a flexible curve to the data and is able to detect non-linear relationships between the variables. When an approximately linear relationship was found, the relationship was summarised with a correlation coefficient.
For correlation of scores between OHIP-Sp5 and either of the two longer versions of the OHIP, we expected to obtain positive correlation values of ‘very large’ (i.e., r > 0.70) size21 as the scores derive from the same instrument, which differs only in the number of items. By contrast, we expected a negative correlation for the OHIP-Sp5/GOHAI-Sp relationship because although these two instruments explore frequency of problems, the scoring system of one is inverse to that of the other. For instance, while for OHIP the higher the score, the worse the perceived OHRQoL (‘bad’ OHRQoL), for GOHAI, the higher the score, the better the self-perception of OHRQoL (‘good’ OHRQoL). Because OHIP-Sp5 and GOHAI-Sp measure the same construct (OHRQoL), we expected ‘large’ (i.e., r > 0.50) coefficients, according to Cohen22. These ‘large’ correlations were also expected for the OHIP-Sp5/OIDP-Sp relationship; however, a positive correlation should be observed here because OHIP and OIDP-Sp are both problem measures. Finally, a ‘large’, but this time negative, correlation should be expected for the OHIP-Sp5/Global Oral Health Assessment relationship. Although both OHIP-Sp5 and the Global Oral Health Assessment measure the same construct, better oral health in the Global Oral Health Assessment is recorded as a higher score, whereas a higher score in the OHIP-Sp5 means worse oral health.
Overall, we expected a pattern of correlations to appear where a larger number of validation instrument items produce higher correlations with OHIP-Sp5. All correlation coefficients were computed using Pearson’s product-moment correlations, except the OHIP-Sp5/Global Oral Health Assessment pairwise, which was computed using Spearman’s rank correlation coefficient.
For all validity analyses, patients with more than one missing OHIP-Sp5 item were excluded from the analysis. For GOHAI-Sp and OIDP-Sp, if more than two items were missing, then the subject’s data were not used. If only one or two items are missing, then imputation is recommended15, 16. When imputation was recommended, missing items for these instruments were imputed using the person median.
All of the statistical procedures were performed using Stata v.14 for Windows (Stata Corp., College Station, TX, USA).
RESULTS
Characterisation of patients
The majority of the patients were females (59.5%) and mean age was 43 years. Almost two-thirds (59.8%) of the participants did not have removable or complete dentures, and only 8.9% of patients had complete dentures.. Patients had an overall OHIP-Sp5 mean value of 3.7 points (Table 3). The 14-item and 49-item versions of the OHIP-Sp had mean scores that were 2.2 and 8.6 times, respectively, higher than that of the OHIP-Sp5. For GOHAI-Sp and OIDP-Sp, overall mean scores were 38.6 and 4.7, respectively. Of all patients, 58.5% self-rated their general oral health as good to excellent.
Table 3.
Descriptive statistics for sociodemographic characteristics and oral health-related quality of life (OHRQoL) measures
| Characteristic | Mean (SD) or Percent* (frequency) |
|---|---|
| Age (years) | 42.8 (12.2) |
| Women | 59.5 (195)* |
| Denture status | |
| No removable or partial denture | 59.8 (195)* |
| Removable partial denture | 11 (36)* |
| Complete dentures | 8.9 (29)* |
| Summary scores of OHRQoL measures | |
| OHIP-Sp49 (points) | 31.9 (35.5) |
| OHIP-Sp14 (points) | 8 (10.2) |
| OHIP-Sp5 (points) | 3.7 (4.0) |
| GOHAI-Sp (points) | 38.6 (7.1) |
| OIDP-Sp (points) | 4.7 (7.6) |
| Global Oral Health Assessment | |
| Excellent | 6.7 (22)* |
| Very good | 18.7 (61)* |
| Good | 33.1 (108)* |
| Fair | 25.1 (82)* |
| Poor | 6.1 (20) |
Indicates absolute frequencies and percentages.
OHIP-Sp5 item characteristics
Mean OHIP-Sp5 item values were highest for ‘difficulty chewing’ and ‘uncomfortable with appearance’, representing the two OHRQoL dimensions Oral Function and Orofacial Appearance, respectively. Painful oral health problems occurred almost with the same frequency. Psychosocial problems, measured using the item ‘difficulty doing usual jobs’ were reported less frequently. The lowest frequency was reported for ‘flavour in food’ – another indicator for the dimension ‘Oral Function’. The minimum response impact (floor effects) and maximum impact (ceiling effect) are shown in Table 4.
Table 4.
Descriptive statistics of Spanish 5-item version of the Oral Health Impact Profile (OHIP-Sp5) items
| OHIP-Sp5 items | Mean (SD) | Median (interquartile range) | Minimum–Maximum (range) | Percentage of participants reporting impact ‘never’ (floor effect) | Percentage of participants reporting impact ‘very often’ (ceiling effect) |
|---|---|---|---|---|---|
| Difficulty chewing | 1.0 (1.1) | 0.0 (2.0) | 0–4 (4) | 41 (135) | 5.2 (17) |
| Painful aching | 0.8 (1.0) | 0.0 (2.0) | 0–4 (4) | 52 (171) | 1.8 (6) |
| Uncomfortable with appearance | 1.0 (1.3) | 0.0 (0.0) | 0–4 (4) | 50 (164) | 10.4 (34) |
| Less flavour in food | 0.2 (0.6) | 0.0 (0.0) | 0–4 (4) | 85 (277) | 0.9 (3) |
| Difficulty doing usual jobs | 0.4 (0.8) | 0.0 (1.0) | 0–4 (4) | 72 (236) | 1.8 (6) |
| OHIP-Sp5 overall score | 3.7 (4.0) | 2.0 (5.0) | 0–19 (19) | n.a. | n.a. |
OHIP-Sp5 overall score is reported as a mean and percents of participants with either ceiling or floor effects cannot be computed for a mean. n.a: non-applicable.
Reliability
Cronbach’s alpha revealed the scale’s internal consistency as ‘good’ (α = 0.83, lower 95% CI limit = 0.80). The estimate did not change substantially when a particular item was deleted from the scale (Table 5).
Table 5.
Internal consistency
| OHIP-Sp5 item | Alpha when item is deleted | Item-rest correlation |
|---|---|---|
| Difficulty chewing | 0.78 | 0.69 |
| Painful aching | 0.79 | 0.65 |
| Uncomfortable with appearance | 0.80 | 0.64 |
| Less flavour in food | 0.82 | 0.59 |
| Difficulty doing usual jobs | 0.78 | 0.70 |
OHIP-Sp5, Spanish 5-item version of the Oral Health Impact Profile.
OHIP-Sp5 items correlated well with each other, as indicated by an average inter-item correlation of 0.53. Individual items also correlated well with the scale. All item-rest correlations (the correlation of each item with the summed index but excluding that item from the index) were 0.59 and higher.
Validity
Validity coefficients for correlation of OHIP-Sp5 summary scores with other OHRQoL measures
When a flexible curve was fitted to OHIP summary scores, all scores correlated linearly (Figure 1, upper panel). A correlation of r = 0.95 was observed between OHIP-Sp5 and both OHIP-Sp49 and OHIP-Sp14. Confidence in the magnitude of this correlation was also very high, as indicated by a very narrow confidence interval around the two-point estimates (Figure 1, lower panel; and Table 6).
Figure 1.
Summary score relationship of the Spanish 5-item version of the Oral Health Impact Profile (OHIP-Sp5) with OHIP-14 and OHIP-49 summary scores, characterised using LOcally WEighted Scatterplot Smoothing (LOWESS) curves (upper panel) and with regression lines and their 95% confidence intervals (lower panel)
Table 6.
Spanish 5-item version of the Oral Health Impact Profile (OHIP-Sp5) scores construct [Spanish 49-item version of the Oral Health Impact Profile (OHIP-Sp49)/Spanish 14-item version of the Oral Health Impact Profile (OHIP-Sp14)] and convergent [Spanish version of the General Oral Health Assessment Index (GOHAI-Sp)/Spanish version of the Oral Impacts on Daily Performances (OIDP-Sp)] validity
| Validity measure | Correlation coefficient* | 95% CI |
|---|---|---|
| OHIP-Sp49score | 0.95 | 0.94 to 0.96 |
| OHIP-Sp14 score | 0.95 | 0.93 to 0.96 |
| OIDP-Sp score | 0.73 | 0.68 to 0.78 |
| GOHAI-Sp score | −0.80† | −0.84 to −0.76 |
| Global Oral Health Assessment | −0.56† | −0.63 to −0.47 |
All correlation coefficients are Pearson’s correlations, except the last, which is Spearman correlation.
For better interpretability, absolute values of the correlation are presented.
The OHIP-Sp5 summary scores also correlated linearly with the summary scores of OIDP-Sp and GOHAI-Sp, the two other OHRQoL instruments. Correlation coefficients among these measures were lower (r=0.73 and r= -0.80) than correlations between OHIP measures (r=0.95). Confidence intervals around correlation coefficients were again very narrow (Figure 2, left and centre panels; and Table 6).
Figure 2.
Summary score relationship of the Spanish 5-item version of the Oral Health Impact Profile (OHIP-Sp5) with Oral Impacts on Daily Performances (OIDP) and General Oral Health Assessment Index (GOHAI) summary scores, characterised using LOcally WEighted Scatterplot Smoothing (LOWESS) curves (upper panel) and with regression lines and their 95% confidence intervals (lower panel)
When OHIP-Sp5 was correlated with the Global Oral Health Assessment, the results changed slightly. The flexible LOWESS line indicated a slightly curved relationship between OHIP-Sp5 and Global Oral Health Assessment. However, a regression line still approximated a linear relationship well with slightly less confidence for patients with high oral health impairment (Figure 2, right panel; and Table 6).
Pattern of validity coefficients
Overall, we observed the expected pattern of absolute magnitude of validity coefficients (Table 5). The highest correlations, ‘very large’ according to Rosenthal21, were observed when OHIP-Sp5 was compared with the same measure of the construct OHRQoL, but containing a larger number of indicators. Comparing five OHRQoL indicators (OHIP-Sp5) with 14 or 49 did not make a difference. The correlation coefficients were almost the same, but the correlation coefficient for OHIP-Sp49 had a minimally smaller confidence interval.
However, when OHIP-Sp5 was compared with other multi-item measures of the same construct OHRQoL, the effect size for the correlation was higher than expected with a still ‘very large’ effect size. As expected, the 8-item instrument had a lower correlation than the 12-item instrument. Finally, when OHIP-Sp5 was compared with a 1-item measure of OHRQoL, the correlation coefficient decreased further, but the effect size was still ‘large’.
DISCUSSION
Overall, when compared with other versions of OHIP-5, the findings of this study were well in line with what was expected from the instrument score. Our study is confirmatory in nature. Study results add to the body of knowledge that OHIP versions are globally robust instruments for measuring OHRQoL in various settings.
Specifically, the OHIP-Sp5 reliability, measured by Cronbach’s alpha, was ‘good’ (α = 0.83), according to Kline19. This finding is consistent with results obtained in Japanese prosthodontic patients (α = 0.81)10, the Swedish general population (α = 0.77)8, the German general population (α = 0.76)9 and English-speaking community participants in the USA (α = 0.75)12. However, in Dutch patients with temporomandibular disorders (TMDs) 7 only an ‘acceptable’ level of reliability (α = 0.67)11 was reached – a value similar to that found for German patients with TMDs (α = 0.65)9. The dimensional structure of OHRQoL, with Oral Function, Orofacial Pain, Orofacial Appearance and Psychosocial Impact as the core components of a patient’s oral health experience, may provide an explanation for the findings6, 7, 23, 24. Prosthodontic patients suffer mainly from tooth loss, an outcome regarded as dentistry’s major or ‘typical’ outcome25. Tooth loss affects all four OHRQoL dimensions. Similarly, participants from the general population, albeit affected to a lesser degree, are also affected in all four dimensions. By contrast, TMD impact is not four-dimensional. These patients are typically not aesthetically impaired. Internal consistency (measured using Cronbach’s alpha) is a measure of the interrelatedness of OHIP-Sp5 items. When all four OHRQoL dimensions are similarly affected, such as in prosthodontic patients and the general population, interrelatedness of the items should be higher, as in populations with fewer dimensions affected, such as in TMD patients – the pattern of Cronbach’s results we described above.
We are able to compare our validity coefficients with other language versions. Some previously validated OHIP-5 versions in Japanese (r = 0.92)10, Swedish (r = 0.92)8 and Dutch (r ≥ 0.90)11 have investigated correlations with OHIP-49. Also, in line with our results is a correlation with the Global Oral Health Assessment in English-speaking participants12. Not all studies have used measures that are comparable with our methodological approach26, 27; nevertheless, the agreement between observed and hypothesised relationships between OHIP-Sp5 scores and clinical measures was interpreted as evidence for score validity.
New in our study are the correlations between OHIP-Sp5 and GOHAI-Sp (r = −0.80), on one hand, as well as between OHIP-Sp5 and OIDP-Sp (r = 0.73) on the other hand. While OHIP-Sp5 was not previously correlated with these dPROMs, OHIP-Sp14 and OHIP-Sp49 were (r = 0.95). Some studies showed correlation coefficients between OHIP-49 and OHIP-5 of 0.92 in Japanese and Swedish dental patients8, 10, which are similar to the values obtained with the shorter 5-item version found in our heterogeneous sample of Spanish-speaking dental patients. Similar values (r=0.90) were observed when correlating OHIP-14 and OHIP-58.
Finally, regarding interpretation of the results obtained for reliability of OHIP-Sp5, the four-dimensional OHRQoL structure with a strong general factor6, 7, 23 also provides the basis for understanding the validity results. In this situation, correlations of OHIP-Sp5 with longer OHIP versions should follow a pattern that higher coefficients are observed for correlations with longer instruments, but representation of the four dimensions in other OHRQoL instruments should also affect correlation results. Because OIDP and GOHAI are also measures of the same construct OHRQoL, although they measure OHRQoL dimension with a different number of items24, large correlations with OHIP-Sp5 should still be observed – a pattern that we identified. Even a one-item indicator of perceived global oral health showed a large correlation with OHIP-Sp5. Altogether, these results suggest that the OHIP-Sp5 had good reliability and validity in dental patients.
Data collection is costly and burdensome. For example, considering broad medical and social science surveys, adding a single item to a large cross-sectional population survey can cost as much as $100,00028. In dental practice settings, patients typically already receive questionnaires of substantial length about their general health, medication, previous treatments, etc. The addition of numerous OHRQoL questions is burdensome. Consequently, a 5-item OHIP provides a pragmatic approach to measuring OHRQoL in all settings. First, it is a short version of one of the most commonly used OHRQoL tools internationally with adequate psychometric properties of its original form, but also its short forms. Above and beyond the 14- and 49-item versions, forms with 3029, 2227, 219, 1930 and 1831 items, and other versions with 14 items32, 33, 34, exist. The OHIP is also, from a methodological point of view, the best investigated questionnaire. For example, score interpretation using normative values35, minimal important difference36 or countable impacts37, have also been investigated.
OHIP-Sp5 has the advantage of being the shortest OHIP instrument. Recently, a Spanish (Chilean) ultrashort OHIP-Sp version with seven items was developed26. When comparing OHIP-7 and OHIP-Sp5, only one item ‘difficulty doing jobs’ is present in both versions. The seven-item version showed ‘excellent’ reliability19, but validity measures were not provided in the study. Compared with that version, OHIP-Sp5 would have the advantage that it has two items fewer, several validated language versions and, most importantly, superior validity by measuring the OHRQoL dimensions Oral Function, Orofacial Pain, Orofacial Appearance and Psychosocial Impact. In a recent systematic review, 53 reported outcome measures for adult dental patients (dPROMs) were identified5. All instruments measured one or more OHRQoL dimensions, but OHIP-5 was the shortest instrument measuring all four dimensions.
In the present study, the response rate was 61% and, therefore, the ability to generalise the results to the target population could be affected by bias. However, for studies such as ours that investigate correlations, non-response bias is typically not a substantial threat because there is often no reason to assume that the correlation between, for example, OHIP-Sp5 and GOHAI scores differ between patients who participated in the study and those who refused. Our sample size is not considered to be small and consequently statistical uncertainty around our validity and reliability results, indicated by narrow confidence intervals, is not large.
We focussed on basic psychometric properties such as reliability and validity of instrument scores. However, other types of validity or reliability and other psychometric properties, such as responsiveness or sensitivity to change, exist but were not investigated. For OHIP in general, information about response format38, retest effects39, administration method effects40, response shift in retrospective OHRQoL assessment41 and recall periods42 is available. We do not consider that there is any reason why such findings should differ across OHIP versions.
CONCLUSIONS
The OHIP-Sp5 questionnaire measures the OHRQoL construct with good reliability and validity. Its ability to capture information is very similar to that of other language versions, supporting OHIP-5 as a suitable instrument not only applicable for Spanish-speaking populations but also in effectively replacing longer versions of OHIP. OHIP-Sp5 is able to characterise OHRQoL with a summary score and also with the dimensions Oral Function, Orofacial Pain, Orofacial Appearance and Psychosocial Impact across all clinical and research settings.
Acknowledgements
We thank Ms. Kathleen Patka, Executive Office and Administrative Specialist, Division of Oral Medicine, School of Dentistry, University of Minnesota, Minneapolis, MN, USA, for proofreading the manuscript.
Research reported in this publication was supported by the National Institute of Dental and Craniofacial Research of the National Institutes of Health under Award Number R01DE022331.
Conflict of interest
None.
REFERENCES
- 1.John MT. Health outcomes reported by dental patients. J Evid Based Dent Pract. 2018;18:332–335. doi: 10.1016/j.jebdp.2018.09.001. [DOI] [PubMed] [Google Scholar]
- 2.Reissmann DR. Dental patient-reported outcome measures are essential for evidence-based prosthetic dentistry. J Evid Based Dent Pract. 2019;19:1–6. doi: 10.1016/j.jebdp.2019.01.003. [DOI] [PubMed] [Google Scholar]
- 3.Listl S. Value-based oral health care: moving forward with dental patient-reported outcomes. J Evid Based Dent Pract. 2019;19:220–224. doi: 10.1016/j.jebdp.2019.101344. [DOI] [PubMed] [Google Scholar]
- 4.Gagliardi DI, Slade GD, Sanders AE. Impact of dental care on oral health-related quality of life and treatment goals among elderly adults. Aust Dent J. 2008;53:26–33. doi: 10.1111/j.1834-7819.2007.00005.x. [DOI] [PubMed] [Google Scholar]
- 5.Mittal H, John MT, Sekulic S, et al. Dental patient-reported outcome measures for adults: a systematic review. J Evid Base Dent Pr. 2018;19:53–70. doi: 10.1016/j.jebdp.2018.10.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.John MT, Reissmann DR, Feuerstahler L, et al. Exploratory factor analysis of the Oral Health Impact Profile. J Oral Rehabil. 2014;41:635–643. doi: 10.1111/joor.12192. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.John MT, Feuerstahler L, Waller N, et al. Confirmatory factor analysis of the Oral Health Impact Profile. J Oral Rehabil. 2014;41:644–652. doi: 10.1111/joor.12191. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Larsson P, John MT, Hakeberg M, et al. General population norms of the Swedish short forms of Oral Health Impact Profile. J Oral Rehabil. 2014;41:275–281. doi: 10.1111/joor.12137. [DOI] [PubMed] [Google Scholar]
- 9.John MT, Miglioretti DL, LeResche L, et al. German short forms of the Oral Health Impact Profile. Community Dent Oral Epidemiol. 2006;34:277–288. doi: 10.1111/j.1600-0528.2006.00279.x. [DOI] [PubMed] [Google Scholar]
- 10.Baba K, Inukai M, John MT. Feasibility of oral health-related quality of life assessment in prosthodontic patients using abbreviated Oral Health Impact Profile questionnaires. J Oral Rehabil. 2008;35:224–228. doi: 10.1111/j.1365-2842.2007.01761.x. [DOI] [PubMed] [Google Scholar]
- 11.Van der Meulen MJ, John MT, Naeije M, et al. Developing abbreviated OHIP versions for use with TMD patients. J Oral Rehabil. 2012;39:18–27. doi: 10.1111/j.1365-2842.2011.02242.x. [DOI] [PubMed] [Google Scholar]
- 12.Naik A, John MT, Kohli N, et al. Validation of the English-language version of 5-item Oral Health Impact Profile. J Prosthodont Res. 2016;60:85–91. doi: 10.1016/j.jpor.2015.12.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Lewis MP, Simons GF, Fennig CD. 20th ed. SIL International, Global Publishing; Dallas, TX: 2017. Ethnologue: Languages of the World; p. 505. [Google Scholar]
- 14.Lopez R, Baelum V. Spanish version of the Oral Health Impact Profile (OHIP-Sp) BMC Oral Health. 2006;6:11. doi: 10.1186/1472-6831-6-11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Adulyanon S, Sheiham A. In: Measuring oral health and quality of life. Slade GD, editor. Department of Dental Ecology, School of Dentistry, University of North Carolina; Chapel Hill, NC: 1997. Oral Impacts on Daily Performances; pp. 151–160. [Google Scholar]
- 16.Atchison KA, Dolan TA. Development of the geriatric oral health assessment index. J Dent Educ. 1990;54:680–687. [PubMed] [Google Scholar]
- 17.Sanchez-Garcia S, Heredia-Ponce E, Juarez-Cedillo T, et al. Psychometric properties of the General Oral Health Assessment Index (GOHAI) and dental status of an elderly Mexican population. J Public Health Dent. 2010;70:300–307. doi: 10.1111/j.1752-7325.2010.00187.x. [DOI] [PubMed] [Google Scholar]
- 18.Cronbach LJ. Coefficient alpha and the internal structure of tests. Psychometrika. 1951;16:297–334. [Google Scholar]
- 19.Kline P. 2nd ed. Routledge; London, UK: 2000. Handbook of Psychological Testing; p. 752. [Google Scholar]
- 20.Montero-Martin J, Bravo-Perez M, Albaladejo-Martinez A, et al. Validation the Oral Health Impact Profile (OHIP-14sp) for adults in Spain. Med Oral Patol Oral Cir Bucal. 2009;14:E44–50. [PubMed] [Google Scholar]
- 21.Rosenthal JA. Qualitative descriptors of strength of association and effect size. J Soc Serv Res. 1996;21:37–59. [Google Scholar]
- 22.Cohen J. 2nd ed. Erlbaum; Hillsdale, NJ: 1988. Statistical Power Analysis for the Behavioral Sciences. [Google Scholar]
- 23.John MT, Rener-Sitar K, Baba K, et al. Patterns of impaired oral health-related quality of life dimensions. J Oral Rehabil. 2016;43:519–527. doi: 10.1111/joor.12396. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.John MT, Reissmann DR, Celebic A, et al. Integration of oral health-related quality of life instruments. J Dent. 2016;53:38–43. doi: 10.1016/j.jdent.2016.06.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.John MT, Reissmann DR, Feuerstahler L, et al. Factor analyses of the Oral Health Impact Profile - overview and studied population. J Prosthodont Res. 2014;58:26–34. doi: 10.1016/j.jpor.2013.11.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Leon S, Correa-Beltran G, De Marchi RJ, et al. Ultra-short version of the oral health impact profile in elderly Chileans. Geriatr Gerontol Int. 2017;17:277–285. doi: 10.1111/ggi.12710. [DOI] [PubMed] [Google Scholar]
- 27.Durham J, Steele JG, Wassell RW, et al. Creating a patient-based condition-specific outcome measure for Temporomandibular Disorders (TMDs): Oral Health Impact Profile for TMDs (OHIP-TMDs) J Oral Rehabil. 2011;38:871–883. doi: 10.1111/j.1365-2842.2011.02233.x. [DOI] [PubMed] [Google Scholar]
- 28.Hays RD, Schalet BD, Spritzer KL, et al. Two-item PROMIS® global physical and mental health scales. J Patient Rep Outcomes. 2017;1:1–5. doi: 10.1186/s41687-017-0003-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Segu M, Collesano V, Lobbia S, et al. Cross-cultural validation of a short form of the Oral Health Impact Profile for temporomandibular disorders. Community Dent Oral Epidemiol. 2005;33:125–130. doi: 10.1111/j.1600-0528.2005.00215.x. [DOI] [PubMed] [Google Scholar]
- 30.Allen F, Locker D. A modified short version of the oral health impact profile for assessing health-related quality of life in edentulous adults. Int J Prosthodont. 2002;15:446–450. [PubMed] [Google Scholar]
- 31.Ide R, Mizoue T, Yamamoto R, et al. Development of a shortened Japanese version of the Oral Health Impact Profile (OHIP) for young and middle-aged adults. Community Dent Health. 2008;25:38–43. [PubMed] [Google Scholar]
- 32.Slade GD. Derivation and validation of a short-form oral health impact profile. Community Dent Oral Epidemiol. 1997;25:284–290. doi: 10.1111/j.1600-0528.1997.tb00941.x. [DOI] [PubMed] [Google Scholar]
- 33.Wong AH, Cheung CS, McGrath C. Developing a short form of Oral Health Impact Profile (OHIP) for dental aesthetics: OHIP-aesthetic. Community Dent Oral Epidemiol. 2007;35:64–72. doi: 10.1111/j.1600-0528.2007.00330.x. [DOI] [PubMed] [Google Scholar]
- 34.Locker D, Allen PF. Developing short-form measures of oral health-related quality of life. J Public Health Dent. 2002;62:13–20. doi: 10.1111/j.1752-7325.2002.tb03415.x. [DOI] [PubMed] [Google Scholar]
- 35.John MT, LeResche L, Koepsell TD, et al. Oral health-related quality of life in Germany. Eur J Oral Sci. 2003;111:483–491. doi: 10.1111/j.0909-8836.2003.00079.x. [DOI] [PubMed] [Google Scholar]
- 36.John MT, Reissmann DR, Szentpetery A, et al. An approach to define clinical significance in prosthodontics. J Prosthodont. 2009;18:455–460. doi: 10.1111/j.1532-849X.2009.00457.x. [DOI] [PubMed] [Google Scholar]
- 37.Reissmann DR, Sierwald I, Heydecke G, et al. Interpreting one oral health impact profile point. Health Qual Life Outcomes. 2013;11:12. doi: 10.1186/1477-7525-11-12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Sierwald I, John MT, Durham J, et al. Validation of the response format of the Oral Health Impact Profile. Eur J Oral Sci. 2011;119:489–496. doi: 10.1111/j.1600-0722.2011.00884.x. [DOI] [PubMed] [Google Scholar]
- 39.John MT, Reissmann DR, Schierz O, et al. No significant retest effects in oral health-related quality of life assessment using the Oral Health Impact Profile. Acta Odontol Scand. 2008;66:135–138. doi: 10.1080/00016350802077835. [DOI] [PubMed] [Google Scholar]
- 40.Reissmann DR, John MT, Schierz O. Influence of administration method on oral health-related quality of life assessment using the Oral Health Impact Profile. Eur J Oral Sci. 2011;119:73–78. doi: 10.1111/j.1600-0722.2010.00805.x. [DOI] [PubMed] [Google Scholar]
- 41.Reissmann DR, Remmler A, John MT, et al. Impact of response shift on the assessment of treatment effects using the Oral Health Impact Profile. Eur J Oral Sci. 2012;120:520–525. doi: 10.1111/j.1600-0722.2012.00998.x. [DOI] [PubMed] [Google Scholar]
- 42.Waller N, John MT, Feuerstahler L, et al. A 7-day recall period for a clinical application of the oral health impact profile questionnaire. Clin Oral Investig. 2016;20:91–99. doi: 10.1007/s00784-015-1484-6. [DOI] [PMC free article] [PubMed] [Google Scholar]