Orthodontic treatment during pregnancy, lactation, and postmenopausal period: a questionnaire development

Abstract This study aimed to develop and validate a self-administered questionnaire in Brazilian Portuguese to verify the level of knowledge of orthodontists in the care of pregnant, lactating, and postmenopausal women, named “Considerations on Orthodontic Treatment during Pregnancy, Lactation, and Postmenopausal Periods.” The development and validation of the questionnaire consisted of the following steps: a) item generation; b) item reduction; c) questionnaire design; and d) validity and reliability tests in a cross-sectional study with 258 orthodontists working in the field from different Brazilian states. A total of 60 orthodontists participated in test-retest over a mean period of 45 days. The preliminary questionnaire consisted of a total of 60 questions. After item reduction, 40 questions were selected for the final version of the questionnaire, with eight questions about pregnant women; six about lactating women; 18 about postmenopausal women, and eight about general knowledge in dentistry. Each item had three response options in the Likert scale format. Face and content validity analysis, reliability assessment through internal consistency (Cronbach’s alpha and McDonald’s omega), and test-retest reliability through the intraclass correlation coefficient (ICC) and Spearman’s correlation coefficient were performed. Face and content validity indicated that the questionnaire was considered valid, objective, and easily understandable. The questionnaire had good internal consistency (Cronbach’s alpha = 0.77; McDonald’s omega = 0.78) and good test-retest reliability (ICC = 0.71; Spearman’s correlation coefficient = 0.51). The questionnaire was considered valid and reliable to assess the level of knowledge of orthodontists in the care of pregnant, lactating, and postmenopausal women.


Introduction
Orthodontics is an expanding area and works extensively with other dental specialties to collaborate with the development or completion of complex oral rehabilitations.][4] During different stages of life, such as during pregnancy, lactation, and postmenopausal period, women undergo major hormonal changes.][7] Estrogen, which is the predominant hormone during a woman's reproductive stage, abruptly decreases its concentration during the postmenopausal period, which can lead to osteoporosis. 8][11] Despite the existing knowledge in the literature about bone and orthodontic tooth movement alterations during pregnancy, lactation, and postmenopausal periods, there are no reports on how orthodontists conduct treatments in these groups of patients.It is of utmost importance that orthodontists be aware of the effects of hormones during the different stages of a woman's life so that their clinical approach may be appropriate according to the demand of the female body, favoring the success of orthodontic treatment. 4,12herefore, we identified the need to develop and validate a relevant assessment instrument to improve and adapt the care provided by orthodontists to women during these periods of their lives.
This study aimed to develop and validate a selfadministered questionnaire in Brazilian Portuguese to verify the level of knowledge of orthodontists in the care of pregnant, lactating, and postmenopausal women, named "Considerations on Orthodontic Treatment during Pregnancy, Lactation, and Postmenopausal Periods."This study hypothesizes that the questionnaire is valid and reliable to verify the level of knowledge of orthodontists in the care of pregnant, lactating, and postmenopausal women.

Methodology Ethical considerations
According to the ethical principles for medical research on human beings -Declaration of Helsinki -the research was approved by the Research Ethics Committee (REC) of the Federal University of Minas Gerais (UFMG) (CAAE: 31864520.9.0000.5149).All the participants of this research have consented to their participation through an informed consent form (ICF).

Instrument development
The instrument was developed according to validated protocols for questionnaire development and validation, and the COSMIN (COnsensus-based Standards for selection of health Measurement INstruments) Box B (reliability) checklist was used to assess the methodological quality of the study design (Figure 1), [13][14][15][16] consisting of the following steps: a) item generation; b) item reduction; c) questionnaire design; and d) survey validity and reliability tests (Figure 2).

Item generation
For item generation, a focus group of experts composed of two PhD professors of pediatric dentistry experienced in the development of questionnaire in the health area, a PhD professor of orthodontics specialized in bone remodeling, mainly in female models, all of them affiliated with the Faculty of Dentistry of the Federal University of Minas Gerais (FAO-UFMG) and, a physician specialized in gynecology and obstetrics, working in a private practice.The expertise group held 19 meetings for developing and adapting the questionnaire items.To address the themes in each section, the expertise group considered the following criteria: the most relevant scientific literature related to the orthodontic treatment during pregnancy, lactation, and postmenopausal periods, the different specific knowledge of each expert, and the physician's experience in clinical practice.The sociodemographic data of the orthodontists were also included in the questionnaire, consisting of: sex, age, state of origin, time of specialization experience, and whether the title of orthodontist was obtained from a public or private faculty or class entities.Finally, the following thematic sections were considered: sociodemographic data, orthodontic treatment in pregnant, lactating, and postmenopausal women, and general clinical issues.
Based on the literature review and on the expertise meetings, a preliminary questionnaire with 60 questions and a five-item response model based on the Likert scale was created and structured as follows: 11 items on knowledge about the care of pregnant women, 11 items on knowledge about the care of lactating women, 27 items on knowledge about the care of postmenopausal women, and 11 questions about general knowledge in dentistry.After the construction and analysis of this first version of the questionnaire, the group of experts decided to add, below each question, a question about the level of relevance of that item: high relevance, medium relevance, or low relevance.At the end of the questionnaire, the respondents could make suggestions, thus giving rise to a second version of the preliminary questionnaire.

Item reduction
Two pilot groups were set up (n = 10 each) for item reduction and questionnaire design.The 20 specialists in the field of orthodontics were selected randomly from different regions of Brazil through professional associations.The first pilot group composed of 10 specialists in orthodontics was invited to remotely answer the preliminary questionnaire in its second version, using Google Forms.The goal was to verify the degree of understanding of orthodontists regarding the questions asked, as well as to assess the level of relevance of each item and welcome possible suggestions for improvement.Based on the analysis of the responses given by the pilot group mentioned above, those items that received two or more low scores in terms of their level of relevance were excluded (Table 1), leading to a reduction in the number of items in the questionnaire from 60 to 40.A question was added below each item, asking whether the item above was clear.Also, if the answer to the question was "no," a justification was requested to allow making a subsequent change to that item.After the first pilot test, the response model changed from five to three items: agree, neither agree nor disagree, and disagree.

Questionnaire design
The second pilot group, composed of other 10 specialists in orthodontics, was invited to participate in the study, also remotely, to assess the objectivity of the questions proposed in the second version of the pilot questionnaire, using Google Forms.After analysis of the answers of the second pilot group by the group of experts, general revisions were made to the statement of each item, misspellings were corrected, and some questions were rephrased, thus establishing the final version of the questionnaire in a self-administered format with 40 items: eight items on knowledge about the care of pregnant women, such as radiographic exams during pregnancy, communication between orthodontists and gynecologists, calcium and vitamin supplementation, and periodicity of prophylactic procedures; six items on the care provided to lactating women regarding hormone replacement, calcium requirements, bone remodeling, and tooth movement; 18 items on knowledge about care in postmenopausal women, such as osteoporosis, physical activity, bone density, hormone replacement, diagnostic tests, and use of bisphosphonates; and eight items on general knowledge in dentistry, such as anamnesis, action of calcium and vitamin D, bone remodeling, and care protocol.The Likert scale was used in the response model, with three response options: "agree," "neither agree nor disagree," and "disagree."The answers were evaluated dichotomously, where 1 indicated a correct answer and 0 indicated a wrong answer.The total sum varied from 0 to 40, and the sums of the thematic sections were as follows: pregnant women (0-8), lactating women (0-6), postmenopausal women (0-18), and general knowledge (0-8).The "neither agree nor disagree" answer contained three items, the "disagree" answer had 11 items, and the "agree" answer included the remainder of the items.The higher the score obtained, the higher the level of knowledge of orthodontists about the subject.The theoretical single domain of the questionnaire was qualitatively assessed through face and content validity.During face validity, it was verified whether the items were able to assess what they were intended to.Content validity verified whether the questionnaire contained questions that covered all aspects of the construct to be assessed. 14Questionnaire validity is defined as the degree to which a measurement quantifies what it intends to measure; in this case, based on the literature and experts.Face and content validity is an essential step for the development of new tests in order to assess the degree of representativeness of the items in the questionnaire standardization process.
In our study, face and content validity was assessed through the application of the preliminary version of the instrument in the pilot and expert groups.In the first step, for face validity, both the suggestions made by both pilot groups of orthodontists, as well as their answers to and suggestions for each question of the questionnaire were analyzed, generating substantial and important changes in the construct.Subsequently, for content validity, a more meticulous and detailed analysis of the question, based on the scientific literature and grammatical rules, was performed by the experts, thus establishing the 40 questions.At this stage, it was possible to verify the degree of understanding of the orthodontists in relation to the formulated items, as well as to assess the level of relevance of each item.

Instrument validity and reliability
For the validity and reliability tests, the number of participants (N) was defined according to the sample calculation method, 14,16,17 in which N should be 5 to 7 times greater than the number of items in the survey (40 items x 5 = 200), including at least 100 participants.Therefore, the survey was applied to a sample group consisting of 258 orthodontists working in the field from different Brazilian states, and who had not participated in the previous pilot groups.The participants were randomly selected from different regions of Brazil through professional associations.Dentists specialized in orthodontics, regardless of age, sex, or training period, were included.Those specialists who were not working in the field were excluded from the study.Professionals were contacted by email via class entities.The email messages contained the link to the questionnaire, information about the study, the study number in the REC, and the ICF.
The questionnaire's reliability was assessed through internal consistency and test-retest.To verify internal consistency, Cronbach's alpha and McDonald's omega were used.After a 4-week interval, 60 orthodontists (approximately 20% of the participants) 18 were invited to participate in the testretest analysis with a mean (range) of 45 (5-69) days from the initial assessments.Test-retest reliability was assessed using Spearman's correlation and the intraclass correlation coefficient (ICC).
The statistical tests run in this study were performed using the Statistical Package for Social Sciences (SPSS for Windows, version 26.0, SPSS Inc. Chicago, USA).The results were expressed as average ± standard deviation (SD).

Results
The validity and reliability of the instrument were assessed in a sample of 258 Brazilian orthodontists.The number of orthodontists was not proportionally distributed by region of the country (Table 2).Table 3 presents the absolute and relative values related to the characteristics of the participants.The mean age of participants was 44.04 (standard deviation of ±11.0) years, minimum of 30 years and maximum of 77 years, and 60.1% of the sample consisted of women.Table 4 shows the average, central tendency, and observed ranges for the total score of the questionnaire, as well as for each group of questions.
During the item reduction process, based on the analysis of the responses given by the first pilot group, 21 items were excluded from the instrument (Table 1), in addition to the generation of a new item -"A special orthodontic care protocol is required for pregnant, lactating, and postmenopausal women" -, which resulted in a 40-item instrument, consisting of eight items on knowledge about care for pregnant women, six items on knowledge about care for lactating women, 18 items on knowledge about care for postmenopausal women, and eight items on general knowledge about dentistry.The participants took approximately 15 minutes to answer the questionnaire.In addition to these important changes, the response model changed from five to three items, consisting of "agree," "neither agree nor disagree," and "disagree" options.Based on the analysis of the responses obtained from the second pilot group, three items were reformulated in the postmenopausal section, as well as two items in the general knowledge section, resulting in the final version of the instrument.Additionally, the order of the sections was changed to give a greater focus on the main objective of the study, establishing the following thematic sections regarding the assessment of the level of knowledge of orthodontists about care: in a) pregnant women -8 items; b) lactating women -6 items; c) postmenopausal women -18 items; and d) general knowledge on dentistry -8 items.
Through qualitative analysis of face validity (by orthodontists) and content validity (by the group of experts), the questionnaire was considered valid, objective, and easy to understand.The questions were clear and the content and answer options were appropriate for what was being evaluated.Table 5 presents the construct plausibility after face and content validity and the pertinent justifications and references of each item of the construct.The questionnaire's reliability was assessed through internal consistency and test-retest reliability.Cronbach's alpha was 0.77 and McDonald's omega was 0.78, demonstrating the good internal consistency of the questionnaire.In the test-retest, the ICC was 0.71, whereas Spearman's correlation coefficient was 0.51,

Discussion
To the best of our k nowledge, this study represents the first approach to the development and validation of an instrument to assess the level of knowledge of orthodontists during the treatment of pregnant, lactating, and postmenopausal women.Validity and reliability are key elements in the evaluation of a measurement instrument. 19In this sense, the main result of this study was the development of a valid and reliable instrument to assess the level of knowledge of orthodontists during the treatment of pregnant, lactating, and postmenopausal women.
There are different instruments in the literature to measure the delivery of health services to the general population. 20Previous studies have indicated the need to raise awareness among general dentists of oral health control in pregnant women, 21 in addition to the need for measures to prevent osteoporosis in the practice of dentistry. 22However, few studies have assessed other aspects of orthodontic treatment, such as the care of pregnant, lactating, and postmenopausal women during oral health treatment. 2,22,23Considering this gap, we verified the need to develop a construct that would assess more specific parameters of the level of knowledge of orthodontists about the care of women in these specific stages of their lives.
A me a su r e me nt i n st r u me nt w it h go o d methodolog ical qualit y is essent ial for the development of any scientific activity. 15,24The first step in the development of this instrument was to carry out a bibliographic survey, looking for a better definition of instrument development and validation, followed by the establishment of a focus group of experts with experience in the subject of this study.4][15][16] Correctly defining the target population to be studied, having a clear and objective language for the target population of the instrument, 19,24,25 and building up the correct methodology for the good development of the study are important and necessary aspects during the process of development and validation of a construct. 15,16The definition of the studied population and the validation of this construct were obtained through a review of the literature, meetings with the group of experts, and access to the pilot groups' suggestions for changes in the instruments.Continuation 7. In order for the orthodontic treatment to be successful in women, it is of the utmost importance that the orthodontist is aware of the hormonal effects during the various stages of female life.
It is necessary that the health professionals gather knowledge about the systemic condition of their patients so that the best diagnosis and prognosis for each specific situation can be established.
4][15][16] Item generation, carried out by the group of experts, resulted in 60 items with five response options.7][28] In line with the literature, two pilot tests were run at different time points and with different purposes in the present study.The first pilot test was performed to assess the level of relevance of each item of the instrument presented in the item reduction phase.A second pilot test was used to verify the objectivity of the proposed questions to determine the questionnaire design after the analysis by the group of experts.In this study, each item of the developed instrument was duly substantiated by the literature, with at least one reference on the topic addressed by the item for content validity.
The final version of the questionnaire was applied to a sample of orthodontists to assess its validity and reliability.A construct is considered valid when it is capable of measuring what is intended to be measured. 20,26,29In this study, the questionnaire was evaluated for face and content validity.Face and content validity tests are meant to verify whether what is being measured with the selected items is really being measured. 30This validity is not determined by statistical analysis, but rather by a qualitative assessment carried out by the group of experts and orthodontists.The evaluations made by both groups showed that the instrument contains clear and objective questions and that the content and answer options are appropriate for what was being evaluated.
Reliability is related to the ability to measure the construct in a consistent, accurate, and stable manner. 16,24Internal consistency is the most widely used indicator of test reliability.In this study, the developed instrument presented satisfactory reliability, which was confirmed by Cronbach's alpha test (0.77) and by McDonald's omega test (0.78). 26,31Although the literature argues that Cronbach's alpha values greater than 0.70 are considered appropriate, alpha values may be lower for scales with few items and in research in which the sample is homogeneous. 32n the other hand, McDonald's omega coefficient is based on a similar model, demonstrating relevant values when this cutoff point is considered. 33ost of the studies test-retest the instrument in an average time of 15-30 days. 14,26,27In this study, the mean application period of the test-retest had a mean of 45 days after the test application.One possible explanation for this delay to obtain the answers was probably due to the fact that the questionnaire was developed, tested, and test-retested during the COVID-19 pandemic.The reliability of a construct tends to decrease when the test-retest application is extended. 34This survey was reapplied with a mean of 45 days, showing a low Spearman's correlation coefficient.Regardless of this fact, our instrument was developed to analyze the knowledge of the orthodontist and there is no consensus in the literature about the test-retest interval for this study profile.Since the instrument developed in this study was an assessment of knowledge rather than of the participant's current health status, such as in the Parental-Caregiver Perceptions Questionnaire (P-CPQ), the slightly longer interval does not affect the current data.Interestingly, in this study, important differences in scores were found between the first and second questionnaire applications.In this sense, participants may have been interested in looking for the right answers after answering the questionnaire.Corroborating this hypothesis, significantly higher average scores of correct answers were observed in the second application of the questionnaire.Anyway, the questionnaire showed good test-retest reliability, demonstrating the instrument's stability.Certainly, in this study, the interval between applications was large, requiring a test-retest with a shorter interval between applications in the future.
According to the proposed sample size calculation method, 14 the sample size for a validation study should be 5 to 7 times greater than the number of items in the instrument, requiring at least 100 participants.The number of participants was determined according to the proposed sample calculation method.Thus, the final number of participants included in this study was 258.The present study included orthodontists from several Brazilian states, thereby improving the reliability of the results.
Although the present paper showed satisfactory results, further studies are needed to assess the structural validity of the instrument and also to evaluate the psychometric properties in distinct populations.The main features of this instrument should be considered for future improvements or as a basis for developing new instruments that aim to assess levels of knowledge in different dental specialties.
Despite reports in the literature on the concern of orthodontists with oral hygiene during orthodontic treatment, 35 few studies have addressed other aspects of orthodontic treatment such as the care of pregnant, lactating, and post-menopausal women. 2,3,4,23A recent study has shown that most patients who seek and accept orthodontic treatment are women. 36This raises the concern about whether orthodontists are able to treat these women in their different stages of life, making it necessary to create a questionnaire to assess the orthodontists' knowledge.To date, there has been no instrument addressing this issue, thus increasing the relevance of the present study.This instrument can be widely used in Brazil, as there are 30,093 registered orthodontists, making orthodontics the dental specialty with the highest number of professionals in the country.Also, the questionnaire will provide access to the knowledge of these specialists about these patients and their treatment.

Conclusion
The instrument "Considerations on Orthodontic Treatment during Pregnancy, Lactation, and Postmenopausal periods" was developed and validated.Th is i n st r ument proved to have satisfactory applicability, validity, and reliability to assess the level of knowledge of orthodontists about the orthodontic treatment of women in their different stages of life.This instrument can be used to encourage orthodontists to seek constant learning about the topic and also encourage individualized, personalized, and comprehensive orthodontic practice concerning pregnant, lactating, and postmenopausal women.Therefore, its use in the field of orthodontics is promising.

Figure 1 .
Figure 1.COSMIN (COnsensus-based Standards for selection of health Measurement INstruments) Box B (reliability) checklist.

Figure 2 .
Figure 2. Flowchart describing the development and validation of the "Considerations on Orthodontic Treatment during Pregnancy, Lactation, and Postmenopausal Periods" (COT-PLP) instrument.

Table 1 .
Results of the pilot test regarding the level of relevance of the items.

Table 2 .
Distribution of the number of orthodontists by Brazilian region

Table 3 .
Descriptive analysis of sociodemographic variables related to orthodontists (n = 258)

Table 5 .
Construct items and answers after face and content validation with justification for the question and references used.