2.1

Study protocol

2.1.1

Pilot trial and sampling procedure

Before starting the main investigation, and following the recommendations of Streiner and Norman , the face and content validities of the QoLFAST-10 scale were empirically checked in a pilot trial that was conducted on a representative sample of patients ( n = 27) that numerically represented 23.68% of the reference population ( n = 114). Although ten (or even fewer) subjects may be enough to assess the precision of instructions, item wording, acceptability of formatting, and ease of administration of a questionnaire , given the population variability in this study , 27 patients ( i.e., around 10 patients per treatment group) were selected for the pilot trial. The participants met the selection criteria established for the main research . Thus, they wore fixed-detachable hybrid implant prostheses ( n = 10; 37.04% of the sample for the pilot trial), cemented implant-supported partial dentures ( n = 9; 33.33%), and screwed implant-supported partial dentures ( n = 8; 29.63%). The comprehensiveness of the QoLFAST-10 index was evaluated by asking the volunteers about the clarity of the items. This guaranteed the validity of the scale for the groups evaluated in the main cross-sectional study .

The reference population was initially composed of 114 patients aged from 40 to 90 years, all of whom had been prosthetically rehabilitated at the Department of Buccofacial Prostheses of the Complutense University of Madrid (U.C.M., Spain) between 2000 and 2015. Subjects were engaged by chronology of past treatment. The inclusion criteria consisted of patients treated with at least: one full-arch hybrid implant prosthesis screwed to 4–6 maxillary and/or interforaminal implants, one cemented partial denture supported by 2–3 implants, or one screwed partial denture supported by 2–3 implants. The exclusion criteria were: patients provided with two different types of the restorations tested (to avoid misinterpretation of the findings), subjects demanding dental treatment, cognitive impairment, motility disorders, implant loss, and/or serious illness .

Patients meeting the selection criteria were invited to take part in this research between January and March of 2016. Those volunteers who agreed to be interviewed for the study were offered a clinical examination free of charge, so that the 107 final participants were scheduled for appointments that were to take place in April of 2016. The patients were distributed into three groups depending on their implant prosthetic rehabilitations: Group 1 (HP; n = 37): hybrid prostheses (control); Group 2 (C-PD, n = 35): cemented partial dentures; and Group 3 (S-PD, n = 35): screwed partial dentures.

This study was conducted in full accordance with the Code of Ethics of the World Medical Association (Declaration of Helsinki), ¹

1 Code of Ethics of the World Medical Association (Declaration of Helsinki). http://www.wma.net/en/30publications/10policies/b3/index.html (accessed 20.08.16).

the Spanish Law 14/2007 for Biomedical Research, ²

2 Spanish Law 14/2007 of July 3rd for Biomedical Research. https://www.boe.es/buscar/doc.php?id=BOE-A-2007-12945 (accessed 20.08.16).

and the Uniform Requirements for manuscripts submitted to Biomedical journals. ³

3 Uniform Requirements for manuscripts submitted to Biomedical journals. http://www.icmje.org/ (accessed 20.08.16).

The approval of the Ethics Committee of the San Carlos University Hospital of Madrid (C.I. 12/242-E, C.I. 12/280-E, and C.I. 14/139-E) was obtained after the ethical board of the health clinic completed an independent review of the research protocol. The study was undertaken with the informed written consent of each participant. The privacy rights of the patients were always observed .

2.1.2

Data gathering

Aided by a trained interviewer who formulated the questions, the participants completed both the QoLFAST-10 and the OHIP-14sp forms, which had previously been validated for other prosthodontic treatments in the same source population .

The QoLFAST-10 questionnaire attempts to cover the whole concept of oral-related well-being that may be assessed when the patient is wearing an implant restoration. The original version of the QoLFAST-10 included the following items: Item 1 (Social repercussion of oral pain), Item 2 (Social repercussion of eating well), Item 3 (Speaking well), Item 4 (Satisfaction with the size, shape, and colour of the prosthetic teeth), Item 5 (Self-confidence when smiling), Item 6 (Development of daily activities), Item 7 (Feeling socially comfortable), Item 8 (Importance of dental revisions), Item 9 (Importance of daily habits of oral hygiene), and Item 10 (Oral hygiene difficulties) . This instrument initially contained the following sub-scales (also called ‘dimensions’ or ‘domains’) and distribution of items ordered from higher to lower factorial weight: Social dimension (Items 7, 2, 1, and 6); Aesthetic dimension (Items 4 and 5); Functional dimension (Items 10 and 3); and Thoughts about health-behavioural habits (Items 9 and 8) .

As detailed in the Results section, an adaptation of the QoLFAST-10 index for implant-supported cemented and screwed FPD wearers was the one used in this study (Appendix A in Supplementary material). In any case, this instrument may be intuitively self-completed as the items’ responses are expressed on a Likert-type scale with proportional codes for the degrees of impact. Hence, the items evaluated as ‘<0’ are considered to have a negative effect, while values of ‘+1’ and ‘+2’ represent the positive side of each item (or at least the absence of a negative impact). The possible responses to the items are the following: ‘strongly disagree’ (score −2), ‘disagree’ (score −1), ‘indecisive’/‘indifferent’/‘neutral’ (score 0), ‘agree’ (score +1), and ‘strongly agree’ (score +2).

Corresponding to the additive method (ADD), the total score of the questionnaire was the sum of the different item scores . Both negative and positive impacts contributed to the total net score in such way that the higher the total score is, the higher the satisfaction of the patient is (meaning that negative or low positive scores indicate poorer self-perceived satisfaction). Therefore, the total score of the QoLFAST-10 ranges between ‘ − 2 × no. of items’ (−20) and ‘+2 × no. of items’ (+20). Correspondingly, the dimensional scores are the sum of the item scores that are included in each sub-scale (ADD method) .

The volunteers also filled out the 14-item Oral Health Impact Profile (OHIP-14sp) form, which had been described in-depth elsewhere . Answering the OHIP-14sp, patients scored in terms of the frequency of appearance, 14 situations of impact that were conceptually divided into seven ‘domains’ or ‘dimensions’ (i.e., Functional limitation, Physical pain, Psychological discomfort, Physical disability, Psychological disability, Social disability, and Handicap) . Frequency in the OHIP was codified using a classic Likert-type scale with five options . The possible impact responses were: ‘hardly ever’ (score +1), ‘occasionally’ (score +2), ‘fairly often’ (score +3), and ‘very often’ (score +4). The ‘never’ response (score 0) revealed the absence of impact. The OHIP-14sp outcome variable ranged from 0 to 56. With this questionnaire, the higher the total score is, the higher the level of negative impact on oral well-being and quality of life is, and, therefore, the lower the satisfaction of the patient is .

Together with the completion of both indices, the patients were also asked about their overall satisfaction with their mouths, which comprised individual assessments of the satisfaction with their oral aesthetics, functionality, and comfort with their prostheses . A visual analogue scale (VAS) was applied for each of the abovementioned areas, so that these perceptions were quantified in a continuous range from 0 to 10 . Subjects could thereby declare themselves to be ‘dissatisfied’, ‘neutral’, or ‘satisfied’, offering values situated left to the midpoint of a 100-mm long line, on the midpoint, or to the right of the midpoint, respectively .

A different investigator conducted each scale. To ensure that the identities of the patients were concealed from the clinic staff, the completed forms were placed in sealed envelopes. The adapted version of the QoLFAST-10 (Appendix A in Supplementary material), the OHIP-14sp, and the VAS evaluations were then linked by means of a unique identification code for each participant .

To capture the clinical modulating factors, subjects were examined by a single researcher using the diagnostic methodology published by the World Health Organization (WHO) .

The study variables were grouped as follows: Group 1: Socio-demographic variables ( gender, age, being partnered, and level of education/schooling ); Group 2: Variables related to the prosthetic rehabilitation ( location, type of antagonist, and status of the prosthesis ); and Group 3: Self-perceived satisfaction with the mouth ( complaints about the mouth , and perception of needing dental treatment ).

2.2

Data analysis

As previously mentioned, the additive method (-ADD) was used for both the QoLFAST-10 and the OHIP-14sp analyses by adding the item codes at the appropriate frequency . The dimensional scores of each questionnaire were obtained in a similar fashion. All of the data collected were processed according to well-established statistical methods used in related research .

Descriptive statistics and percentages for qualitative and categorical variables were calculated . The psychometric capacity of the QoLFAST-10 questionnaire ( i.e., reliability and validity) was investigated in implant-supported cemented and screwed FPD wearers. On the one hand, as each item measured different aspects of the same attribute, the reliability was assessed by examining the internal consistency of the scales through the use of the Cronbach’s α value, the α value if an item was deleted, the inter-item correlation, and the item-total correlation . On the other hand, different types of validity were tested:

(a) The face and content validities (or the extent to which a measure represents all facets of a given construct) were verified in the pilot trial because the patients reported no difficulties in understanding the items and did not mention any situation of impact that had not been included in the questionnaire .

(b) The construct validity of the QoLFAST-10 (or the extent to which the OHRQoL was actually recorded with this scale) was examined using the factor analysis (a data reduction technique that allows homogeneous subgroups of variables to be found), and the convergent validity (which measures how closely the new instrument is related to other variables and measures of the same construct to which it should be associated) . As regards the factor analysis, the principal components’ analysis (PCA) was applied along with the rotation method: the Varimax plus Kaiser normalization was chosen to extract the underlying domains of the prosthetic well-being construct . Afterwards, the Bartlett’s Sphericity and the Kaiser-Meyer-Olkin (KMO) tests, which are measures of sampling adequacy, were run to detect the factorial structure of the QoLFAST-10 in C-PD and S-PD wearers . Factors with an eigenvalue of less than one were disregarded to avoid distortion . The items were assigned to the rotated factors (or dimensions) when they had a loading of 0.5 or greater on a single factor . To establish the degree of convergent validity, the QoLFAST-10 total and sub-scale scores were correlated to the total score of the OHIP-14sp with the Spearman’s rank correlation test .

(c) The criterion validity of the QoLFAST-10 indicator (which measures how well the test predicts the OHRQoL based on information obtained from other variables) was analysed by contrasting the total QoLFAST-10 and OHIP-14sp scores with the VAS punctuations using non-parametric probes, since the Kolmogorov-Smirnov test did not assume a normal distribution of the QoLFAST-10 outcome variable in the prosthodontic groups evaluated ( p < 0.01). Therefore, the Kruskal-Wallis test was applied for variables with three or more categories, while the Mann-Whitney U test was run for variables with two categories, and for pair-wise comparisons .

(d) With the aim to investigate the discriminant validity, the total and dimensional scores of the QoLFAST-10 were compared with the total score of the OHIP-14sp (and vice versa ) among the prosthodontic groups. The Kruskal-Wallis test was used to assess the differences among the three prosthodontic groups, and the Mann-Whitney U test was chosen for post-hoc comparisons .

After evaluating the psychometric characteristics of the QoLFAST-10 questionnaire for C-PD and S-PD wearers, the modulating factors of the prosthetic well-being construct were explored. The Kruskal-Wallis and the Mann-Whitney U tests were thus run to evaluate the influence of the study variables on the impact scores of both the QoLFAST-10 and the OHIP-14sp .

Data were processed using the Statistical Package for the Social Sciences (software v.22) (SPSS/PC+, Inc.; Chicago, IL, USA) taking in advance the cut-off level for statistical significance at α = 0.05 .

2.1

Study protocol

2.1.1

Pilot trial and sampling procedure

Before starting the main investigation, and following the recommendations of Streiner and Norman , the face and content validities of the QoLFAST-10 scale were empirically checked in a pilot trial that was conducted on a representative sample of patients ( n = 27) that numerically represented 23.68% of the reference population ( n = 114). Although ten (or even fewer) subjects may be enough to assess the precision of instructions, item wording, acceptability of formatting, and ease of administration of a questionnaire , given the population variability in this study , 27 patients ( i.e., around 10 patients per treatment group) were selected for the pilot trial. The participants met the selection criteria established for the main research . Thus, they wore fixed-detachable hybrid implant prostheses ( n = 10; 37.04% of the sample for the pilot trial), cemented implant-supported partial dentures ( n = 9; 33.33%), and screwed implant-supported partial dentures ( n = 8; 29.63%). The comprehensiveness of the QoLFAST-10 index was evaluated by asking the volunteers about the clarity of the items. This guaranteed the validity of the scale for the groups evaluated in the main cross-sectional study .

The reference population was initially composed of 114 patients aged from 40 to 90 years, all of whom had been prosthetically rehabilitated at the Department of Buccofacial Prostheses of the Complutense University of Madrid (U.C.M., Spain) between 2000 and 2015. Subjects were engaged by chronology of past treatment. The inclusion criteria consisted of patients treated with at least: one full-arch hybrid implant prosthesis screwed to 4–6 maxillary and/or interforaminal implants, one cemented partial denture supported by 2–3 implants, or one screwed partial denture supported by 2–3 implants. The exclusion criteria were: patients provided with two different types of the restorations tested (to avoid misinterpretation of the findings), subjects demanding dental treatment, cognitive impairment, motility disorders, implant loss, and/or serious illness .

Patients meeting the selection criteria were invited to take part in this research between January and March of 2016. Those volunteers who agreed to be interviewed for the study were offered a clinical examination free of charge, so that the 107 final participants were scheduled for appointments that were to take place in April of 2016. The patients were distributed into three groups depending on their implant prosthetic rehabilitations: Group 1 (HP; n = 37): hybrid prostheses (control); Group 2 (C-PD, n = 35): cemented partial dentures; and Group 3 (S-PD, n = 35): screwed partial dentures.

This study was conducted in full accordance with the Code of Ethics of the World Medical Association (Declaration of Helsinki), ¹

1 Code of Ethics of the World Medical Association (Declaration of Helsinki). http://www.wma.net/en/30publications/10policies/b3/index.html (accessed 20.08.16).

the Spanish Law 14/2007 for Biomedical Research, ²

2 Spanish Law 14/2007 of July 3rd for Biomedical Research. https://www.boe.es/buscar/doc.php?id=BOE-A-2007-12945 (accessed 20.08.16).

and the Uniform Requirements for manuscripts submitted to Biomedical journals. ³

3 Uniform Requirements for manuscripts submitted to Biomedical journals. http://www.icmje.org/ (accessed 20.08.16).

The approval of the Ethics Committee of the San Carlos University Hospital of Madrid (C.I. 12/242-E, C.I. 12/280-E, and C.I. 14/139-E) was obtained after the ethical board of the health clinic completed an independent review of the research protocol. The study was undertaken with the informed written consent of each participant. The privacy rights of the patients were always observed .

2.1.2

Data gathering

Aided by a trained interviewer who formulated the questions, the participants completed both the QoLFAST-10 and the OHIP-14sp forms, which had previously been validated for other prosthodontic treatments in the same source population .

The QoLFAST-10 questionnaire attempts to cover the whole concept of oral-related well-being that may be assessed when the patient is wearing an implant restoration. The original version of the QoLFAST-10 included the following items: Item 1 (Social repercussion of oral pain), Item 2 (Social repercussion of eating well), Item 3 (Speaking well), Item 4 (Satisfaction with the size, shape, and colour of the prosthetic teeth), Item 5 (Self-confidence when smiling), Item 6 (Development of daily activities), Item 7 (Feeling socially comfortable), Item 8 (Importance of dental revisions), Item 9 (Importance of daily habits of oral hygiene), and Item 10 (Oral hygiene difficulties) . This instrument initially contained the following sub-scales (also called ‘dimensions’ or ‘domains’) and distribution of items ordered from higher to lower factorial weight: Social dimension (Items 7, 2, 1, and 6); Aesthetic dimension (Items 4 and 5); Functional dimension (Items 10 and 3); and Thoughts about health-behavioural habits (Items 9 and 8) .

As detailed in the Results section, an adaptation of the QoLFAST-10 index for implant-supported cemented and screwed FPD wearers was the one used in this study (Appendix A in Supplementary material). In any case, this instrument may be intuitively self-completed as the items’ responses are expressed on a Likert-type scale with proportional codes for the degrees of impact. Hence, the items evaluated as ‘<0’ are considered to have a negative effect, while values of ‘+1’ and ‘+2’ represent the positive side of each item (or at least the absence of a negative impact). The possible responses to the items are the following: ‘strongly disagree’ (score −2), ‘disagree’ (score −1), ‘indecisive’/‘indifferent’/‘neutral’ (score 0), ‘agree’ (score +1), and ‘strongly agree’ (score +2).

Corresponding to the additive method (ADD), the total score of the questionnaire was the sum of the different item scores . Both negative and positive impacts contributed to the total net score in such way that the higher the total score is, the higher the satisfaction of the patient is (meaning that negative or low positive scores indicate poorer self-perceived satisfaction). Therefore, the total score of the QoLFAST-10 ranges between ‘ − 2 × no. of items’ (−20) and ‘+2 × no. of items’ (+20). Correspondingly, the dimensional scores are the sum of the item scores that are included in each sub-scale (ADD method) .

The volunteers also filled out the 14-item Oral Health Impact Profile (OHIP-14sp) form, which had been described in-depth elsewhere . Answering the OHIP-14sp, patients scored in terms of the frequency of appearance, 14 situations of impact that were conceptually divided into seven ‘domains’ or ‘dimensions’ (i.e., Functional limitation, Physical pain, Psychological discomfort, Physical disability, Psychological disability, Social disability, and Handicap) . Frequency in the OHIP was codified using a classic Likert-type scale with five options . The possible impact responses were: ‘hardly ever’ (score +1), ‘occasionally’ (score +2), ‘fairly often’ (score +3), and ‘very often’ (score +4). The ‘never’ response (score 0) revealed the absence of impact. The OHIP-14sp outcome variable ranged from 0 to 56. With this questionnaire, the higher the total score is, the higher the level of negative impact on oral well-being and quality of life is, and, therefore, the lower the satisfaction of the patient is .

Together with the completion of both indices, the patients were also asked about their overall satisfaction with their mouths, which comprised individual assessments of the satisfaction with their oral aesthetics, functionality, and comfort with their prostheses . A visual analogue scale (VAS) was applied for each of the abovementioned areas, so that these perceptions were quantified in a continuous range from 0 to 10 . Subjects could thereby declare themselves to be ‘dissatisfied’, ‘neutral’, or ‘satisfied’, offering values situated left to the midpoint of a 100-mm long line, on the midpoint, or to the right of the midpoint, respectively .

A different investigator conducted each scale. To ensure that the identities of the patients were concealed from the clinic staff, the completed forms were placed in sealed envelopes. The adapted version of the QoLFAST-10 (Appendix A in Supplementary material), the OHIP-14sp, and the VAS evaluations were then linked by means of a unique identification code for each participant .

To capture the clinical modulating factors, subjects were examined by a single researcher using the diagnostic methodology published by the World Health Organization (WHO) .

The study variables were grouped as follows: Group 1: Socio-demographic variables ( gender, age, being partnered, and level of education/schooling ); Group 2: Variables related to the prosthetic rehabilitation ( location, type of antagonist, and status of the prosthesis ); and Group 3: Self-perceived satisfaction with the mouth ( complaints about the mouth , and perception of needing dental treatment ).

2.2

Data analysis

As previously mentioned, the additive method (-ADD) was used for both the QoLFAST-10 and the OHIP-14sp analyses by adding the item codes at the appropriate frequency . The dimensional scores of each questionnaire were obtained in a similar fashion. All of the data collected were processed according to well-established statistical methods used in related research .

Descriptive statistics and percentages for qualitative and categorical variables were calculated . The psychometric capacity of the QoLFAST-10 questionnaire ( i.e., reliability and validity) was investigated in implant-supported cemented and screwed FPD wearers. On the one hand, as each item measured different aspects of the same attribute, the reliability was assessed by examining the internal consistency of the scales through the use of the Cronbach’s α value, the α value if an item was deleted, the inter-item correlation, and the item-total correlation . On the other hand, different types of validity were tested:

(a) The face and content validities (or the extent to which a measure represents all facets of a given construct) were verified in the pilot trial because the patients reported no difficulties in understanding the items and did not mention any situation of impact that had not been included in the questionnaire .

(b) The construct validity of the QoLFAST-10 (or the extent to which the OHRQoL was actually recorded with this scale) was examined using the factor analysis (a data reduction technique that allows homogeneous subgroups of variables to be found), and the convergent validity (which measures how closely the new instrument is related to other variables and measures of the same construct to which it should be associated) . As regards the factor analysis, the principal components’ analysis (PCA) was applied along with the rotation method: the Varimax plus Kaiser normalization was chosen to extract the underlying domains of the prosthetic well-being construct . Afterwards, the Bartlett’s Sphericity and the Kaiser-Meyer-Olkin (KMO) tests, which are measures of sampling adequacy, were run to detect the factorial structure of the QoLFAST-10 in C-PD and S-PD wearers . Factors with an eigenvalue of less than one were disregarded to avoid distortion . The items were assigned to the rotated factors (or dimensions) when they had a loading of 0.5 or greater on a single factor . To establish the degree of convergent validity, the QoLFAST-10 total and sub-scale scores were correlated to the total score of the OHIP-14sp with the Spearman’s rank correlation test .

(c) The criterion validity of the QoLFAST-10 indicator (which measures how well the test predicts the OHRQoL based on information obtained from other variables) was analysed by contrasting the total QoLFAST-10 and OHIP-14sp scores with the VAS punctuations using non-parametric probes, since the Kolmogorov-Smirnov test did not assume a normal distribution of the QoLFAST-10 outcome variable in the prosthodontic groups evaluated ( p < 0.01). Therefore, the Kruskal-Wallis test was applied for variables with three or more categories, while the Mann-Whitney U test was run for variables with two categories, and for pair-wise comparisons .

(d) With the aim to investigate the discriminant validity, the total and dimensional scores of the QoLFAST-10 were compared with the total score of the OHIP-14sp (and vice versa ) among the prosthodontic groups. The Kruskal-Wallis test was used to assess the differences among the three prosthodontic groups, and the Mann-Whitney U test was chosen for post-hoc comparisons .

After evaluating the psychometric characteristics of the QoLFAST-10 questionnaire for C-PD and S-PD wearers, the modulating factors of the prosthetic well-being construct were explored. The Kruskal-Wallis and the Mann-Whitney U tests were thus run to evaluate the influence of the study variables on the impact scores of both the QoLFAST-10 and the OHIP-14sp .

Data were processed using the Statistical Package for the Social Sciences (software v.22) (SPSS/PC+, Inc.; Chicago, IL, USA) taking in advance the cut-off level for statistical significance at α = 0.05 .