ABSTRACT
Objective:
In the present study, it was aimed to adapt the fetal health anxiety inventory (FHAI) into Turkish and to analyze the validity and reliability of the scale among pregnant women.
Methods:
Explanatory factor analysis (EFA) was applied to 370 pregnant women in Sample I and confirmatory factor analysis (CFA) was applied to 200 pregnant women in Sample II. The Prenatal Distress Questionnaire (NuPDQ) was used to test criterion-related validity of the FHAI. The reliability of the inventory was examined with Cronbach’s alpha reliability coefficient, item-total score correlation coefficient and test-retest analysis.
Results:
As a result of EFA applied to Sample I, it was determined that the 14-item FHAI covered a single factor, and the scale demonstrated good fit indices (χ2/standard deviation =3.148, comparative fit index =0.907, standardized root mean squared residual =0.000, root mean square error of approximation =0.089, and p value =0.000) as a result of the CFA applied to Sample II. A statistically significant positive correlation was found between the FHAI and NuPDQ (r=0.851, p<0.01). Cronbach’s alpha internal consistency coefficient of the inventory was 0.85, and item-total score correlation coefficients were found to range between r=0.34- 0.59 (p<0.001). In the test-retest analysis, a statistically significant and positive correlation was found between the total scores of the inventory obtained in two different applications (r=0.568, p=0.001).
Conclusion:
The current study supported the use of 14-item FHAI as a valid and reliable tool to measure fetal health anxiety of Turkish pregnant women.
Introduction
Stress and mental illnesses experienced during pregnancy may have negative effects on both the mother and the developing infant (1-4). The physiological changes required for the growth of fetus and development of fetal health as well as response to these physiological changes take place throughout pregnancy, and these changes are accompanied with dread and uncertainty. The well-being of the pregnant woman is very important both for herself and for the health of the developing fetus. If the expectant mother cannot cope with her anxiety, the risk of mental and psychological disorders increases and the anxiety negatively affects the mechanisms that enable the adaptation process to take place in a healthy way (1,2).
There are many studies in the literature showing negative birth outcomes such as low birth weight in infants, small head circumference, which is an indicator of brain development, and preterm labor associated with the stress experienced during pregnancy (5-7). At the same time, there are findings showing that the preterm is negatively affected in neurodevelopmental, emotional, and behavioral areas (8). Concerns over the health of the mother’s growing fetus are referred to as fetal health anxiety (4,9). Fetal health is affected by maternal, fetal, placental, and external factors. Early diagnosis and treatment of problems that may adversely affect fetal health minimize fetal mortality and morbidity (10). Pregnant women should be examined in order to determine fetal health concerns during pregnancy. Using a qualitative methodology, Harpel (4) investigated the influence of ultrasonography on the experiences of fetal health concerns among 30 pregnant women. According to the findings, ultrasonography can help women feel less anxious about their fetuse’s health, especially when they can see the image and hear positive comments (4). At the same time, it is stated in the literature that the knowledge that the fetus is healthy following an ultrasound reduces the anxiety levels of parents (11,12).
Although there are many measurement tools and studies that analyze the mother’s anxiety level during pregnancy (13-16), the fetal health anxiety cannot be determined because there is no measurement tool that determines it. The fact that fetal issues are widely acknowledged emphasizes the necessity to address the measuring techniques that may be employed to test for this issue. In actuality, it’s crucial to examine whether the measuring techniques to be employed for this goal are appropriate for various cultural frameworks. By adapting the Fetal Health Anxiety Inventory (FHAI) to Turkish in terms of determining fetal health anxiety, creating strategies to eliminate anxiety, and defining it more specifically and objectively, the current study aimed to test the validity and reliability of the FHAI in Turkish population.
Methods
Study Design and Participants
The current research was designed in methodological type. The study involved expecting mothers who were enrolled in a public hospital’s pregnancy course in eastern Turkey. Records showed that the researchers contacted pregnant women who satisfied the inclusion requirements and told them about the study. Women who accepted to take part in the study were asked to complete a web-based survey. The study questionnaires were created with the help of the Google Forms program (https://docs.google.com/forms), and links to the surveys were distributed to the expectant mothers via social media such as WhatsApp, and Facebook Messenger. On the first page of the online survey, there was information on the study and a consent form for participants. The data collection phase was completed by filling out the questionnaires, which took approximately 5-10 minutes.
The sample size for the current investigation was decided upon using the standards proposed by Comrey and Lee (16). Comrey and Lee (16) advised the following sample size guidelines for factor analysis: Very poor was defined as 50, poor as 100, moderate as 200, good as 300, very good as 500, and excellent as 1,000 or above. Participants in the research included 570 pregnant women. Two separate sample groups were subjected to confirmatory factor analysis (CFA) and exploratory factor analysis (EFA). Three hundred seventy pregnant women from sample I were given EFA, while 200 pregnant women from sample II were given CFA. To determine if the sample size was enough, a post hoc power analysis was carried out. Web-based and publicly available statistical software OpenEpi version 3.01 was used for power analysis (17). The study’s power was determined to be 95% for 570 pregnants that were part of the sample, with a 5% margin of error, bipolar significance level, and a 95% confidence interval.
The data from Sample I was collected between the dates of August and October, 2021, and the data from Sample II was collected in December, 2021 (Figure 1). Pregnant women who met the inclusion criteria were randomly selected from the records. The inclusion criteria were being a healthy pregnant woman who was older than 18 years of age, without any psychiatric disease or depressive symptoms.
Data Collection Tools
Personal Information Form
It is a form consisting of questions to determine the sociodemographic characteristics of pregnant women (such as age, educational level, employment status) and obstetric characteristics (parity, the status of having a planned/desired pregnancy, trimester) (1,4,10).
Prenatal Distress Questionnaire (NuPDQ)
It was created by Yali and Lobel (18) to gauge pregnant women’s prenatal discomfort levels. A 17-item version was developed by Lobel (19) as a result of revision studies, and by Yuksel et al. (20), its Turkish validity and reliability study was established. The questionnaire determines the physical and emotional symptoms that may occur during pregnancy, the levels of anxiety and concern experienced by women in matters related to motherhood, body image, and pregnancy. The questionnaire consists of four subscales: “Physical and Social Changes related to Pregnancy, Concerns regarding the Baby and Labor”, “Concerns regarding the Quality of Healthcare and Health Status”, “Concerns regarding the Baby-care and Postpartum Life”, and “Financial Concerns”. A minimum of 0 points and a maximum of 34 points can be obtained from the entire questionnaire. An increase in the scores of the questionnaire and its subscales indicates that the level of distress perceived by pregnant women increases. After doing their research, Yuksel et al. (20) discovered that the questionnaire’s internal consistency coefficient was 0.85. The scale’s internal consistency coefficient according to Cronbach’s alpha was estimated to be 0.77 in the current investigation.
Fetal Health Anxiety Inventory (FHAI)
Reiser and Wright (21) created the FHAI, and this study examined its validity and reliability. The survey gauges pregnant women’s worry over the fetus’s well-being. There are no subscales in the FHAI and it consists of 14 items. Each item consists of 4 statements that best deal with the experiences of the pregnant women in the past weeks. Items on the inventory are scored between 0 (no symptoms) and 3 (severe symptoms), and the sum of the items gives the total score of fetal health anxiety (0-42), while the higher the score, the higher the fetal health anxiety level (21) (Appendix 1).
Process of the Cultural Adaptation
The main task of cultural adaptation was to translate the FHAI into Turkish. The stages of language validity, content validity, and pilot implementation made up the cultural adaptation phase.
Language validity: In the process of adapting the FHAI to Turkish, first of all, permission was obtained from Wright, who developed the scale, via e-mail. The scale was translated into Turkish during the language validation phase of the test by two faculty members and two independent linguists (lecturers in the field of midwifery, obstetrics, and gynecology nursing). The researchers looked over the translations and developed the Turkish version that most accurately reflected each item. Two separate, qualified linguists translated the generated Turkish text back into English and checked for compatibility between the two languages. It was discovered that the back translation form and the original inventory were identical.
Content Validity Index for Items (I-CVI)
Content validity was performed for language and content control of the items of the inventory (22,23). There are two most common methods used for content validity. One of them is the Lawshe technique and the other is the Davis technique. Davis technique was used for the content validity index in our study. In the Davis technique, experts are asked to evaluate their opinions with one of four degrees. In this scoring, the statements are 1 point for “not appropriate”, 2 points for “somewhat appropriate, the item needs correction”, 3 points for “quite appropriate but minor changes are necessary”, and 4 points for “very appropriate” (22). To determine the content validity of the scale, five faculty members, four in midwifery and one in obstetrics and gynecology, were asked to rate each item of the scale from 1 to 4. After examining the average scores given by the experts for each item of the scale, it is recommended to completely remove or rearrange the items that are at least below the agreement limit or least compatible (24). The “content validity index (CVI)” of the item is obtained by dividing the number of experts marking the options in this technique by the total number of experts. As a result of the evaluation of expert opinions using Davis method; while the statements that the experts found very appropriate were accepted without changing, the statements that the experts did not find appropriate or wanted to be corrected were revised and corrected. A CVI greater than 0.80 means that the content validity of the scale is statistically significant. As a result of the expert evaluations of the scale; the CVI calculated with the CVI formula was found to be quite high (CVI =0.95, Table 1). It was seen that the scores obtained from the experts were not statistically different and there was a consistency between the experts. The level of agreement with the expert opinions was examined with the Kendall W analysis, which was a non-parametric test (25). In line with expert opinions, the draft version of the FHAI inventory was completed.
Pilot implementation: In the next stage, the trial Turkish form of the inventory was applied to 10 pregnant women and they were asked to identify the unclear expressions. The findings obtained were not included in the results of the present research. There were no misunderstood items in the pilot implementation, and the Turkish version of the FHAI was completed.
Psychometric Testing of the FHAI
During the adaptation to Turkish of the FHAI validity-reliability analyses were performed for psychometric analysis.
Validity
Exploratory Factor Analysis
First, the applicability of the dataset was assessed in order to assess the construct validity of the FHAI. Kaiser-Meyer-Olkin (KMO) and Barlett’s test of sphericity were applied for this. The KMO must exceed 0.60 in order for the data to be appropriate for factor analysis, and Bartlett’s test of sphericity results must be statistically significant (26). For the inventory’s construct validity, explanatory and confirmatory factor analyses were carried out.
EFA was used to identify the scale’s fundamental elements. A high variance ratio as a result of EFA indicates that the factor structure power of the scale is high. A total explained variance above 50% indicates that there is strong construct validity (27). It is advised to remove goods with factor loading below 0.30 from the inventory as this is the maximum value for factor loading (28). EFA was used to analyze the data from Sample I (n=500).
Confirmatory Factor Analysis
By comparing the results of one or more measurement tools that are assumed to measure the same feature with the score of the measurement tool that is meant to assess the desired feature as a standard, criterion-related validity/concurrent validity is produced. Similar results show that the measurement tool has criterion-related validity when the measurement tool’s validity has been studied and compared to the measurement tool of which validity has already been established (29). Pregnant women were subjected to the NuPDQ in order to assess the FHAI’s criterion-related validity.
CFA is an analysis in which a previously defined and limited structure is tested to whether it is validated as a model (30). CFA was applied to test the structure obtained after EFA. For CFA studies, the FHAI was applied to sample II and the obtained data was analyzed. Multiple fit indices were used for CFA and chi-square goodness, comparative fit index (CFI), Standardized Root Mean Square Residual (SRMR), Root Mean Square Error of Approximation (RMSEA), and p value for testing the null hypothesis’s fit indices were examined. In the fit indices, c2/df <5, 0.90≤ CFI, 0.05< SRMR ≤0.10, 0.05≤ RMSEA ≤0.10, and p value <0.05 were regarded as acceptable criteria (31).
Reliability
Cronbach’s Alpha Reliability Coefficient
The reliability of the FHAI was measured by calculating the Cronbach’s alpha internal consistency coefficient, determining item-total score correlations, and test-retest reliability. Internal consistency increases as the coefficient approaches 1, and in order to say that a scale is reliable, the calculated coefficient must be at least 0.70 (32).
Item-total Score Correlation Coefficient
The item-total score correlation coefficients were looked at in order to look at the link between the scores acquired from the FHAI’s items and the overall score of the inventory. The item-total score correlation coefficient provides details on how the assessment tool’s items relate to one another. It is advised that the acceptable item selection coefficient be higher than 0.20 (33).
Test‐retest Analysis
In order to determine the invariance of the FHAI over time, the consistency of the responses of the individuals to the items of the measurement tool by applying the measurement tool to the same individuals at different times shows the invariance of the measurement tool against time. In the literature, it is stated that at least 30 participants should be reached for the test-retest (34), and the scale was applied to 30 women for the second time to be retested 15 days later. The correlation coefficient between the values obtained at the end of the two applications gives the reliability coefficient of the scale. As this value takes values between 0 and 1 and gets closer to 1, the reliability of the correlation value increases (35).
Statistical Analysis
The SPSS 25.0 (SPSS Inc., Chicago, IL, USA) and AMOS 24 (Multivariate Software, Inc., Los Angeles, CA, USA) were used to analyze the study’s data (36). The Kolmogorov-Smirnov test was performed to examine the results of all scale values utilized in this study in order to verify the normalcy assumptions. Due to the data’s normal distribution, parametric tests were applied. Descriptive statistics including frequency, percentage, mean, and standard deviation were used to describe the characteristics of the pregnant women. To examine the content validity of the inventory, Kendall’s W-test was applied. The data’s acceptability for factor analysis was assessed using KMO analysis, and the sample size’s appropriateness was determined using Bartlett’s test of Sphericity. EFA and the Promax Rotation technique were used to investigate the factor structure of the inventory. To validate the structure of the inventory discovered by factor analysis, CFA was carried out. The Pearson product-moment correlation analysis was used to look at the correlation between the FHAI and NuPDQ as part of the criterion-related validity/concurrent validity investigation of the inventory. In the reliability analyses of the inventory, the item analysis and test-retest analysis were investigated using Pearson correlation analysis, and the internal consistency of the inventory was examined with Cronbach’s alpha reliability coefficient. Explanatory factor analysis produced the main findings. Additionally, analyses of the content validity, confirmatory factor analysis, criterion-related validity, and reliability were carried out.
Ethical Issues
Wright, who created the scale, was contacted by email to provide his consent to translate the FHAI into Turkish. Additionally, approval from the Scientific Research and Publication Ethics Committee of the İnönü University (decision number: 2020/915) was acquired prior to data collection. On the first page of the online questionnaire, which served as the study’s consent form, all pregnant women who agreed to participate in it were informed about the research before it began. Included were those who freely took part in the study.
Results
A total of 2% of pregnant women with high-risk pregnancies (such as cardiac conditions and gestational diabetes) were cut from the study’s sample of EFA (n=377) due to inclusion requirements. Of pregnant women with hazardous pregnancies (including placenta previa, intrauterine growth retardation, and high blood pressure) 4% were eliminated from the research after the target sample (n=208) was checked for CFA inclusion requirements. Two hundred pregnant women in Sample II of the study’s CFA and 370 pregnant women in Sample I of its EFA were both included in the analysis.
Participants’ Characteristics
The mean gestational week of Sample I included in the study was 36.72±4.53, 64.3% of them were between the ages of 18 and 30, 36.5% were university graduates, 78.6% were unemployed, 84.1% had the moderate level of income, 88.4% had planned pregnancy, 97.6% had the desired pregnancy, 95.4% were in the third trimester, and 62.2% were multiparas. The mean gestational week of Sample II was 30.62±9.53, 65.5% of them were between the ages of 18 and 30, 50.0% were university graduates, 65.5% were unemployed, 73.5% had moderate level of income, 82.5% had planned pregnancy, 91.5% had desired pregnancy, 74.5% were in the third trimester, and 68.5% were multiparas (Table 2).
Validity
Exploratory Factor Analysis
The items evaluated as a result of the expert views reviewed by the Kendall W analysis were not significantly different from one another for the content validity of the scale (Kendall W =0.851; p>0.05), and the expert opinions were consistent. Following the application of EFA to Sample I, the KMO value of FHAI was 0.851 and the results of Barlett’s test of sphericity were X2 =1955.003 and p=0.001, respectively. According to the results of the investigation, the sample size was suitable for factor analysis. According to the EFA findings, the sole factor with an eigenvalue above 1 was found for 14 FHAI components. Between 0.4 and 0.69 factor loadings accounted for 37.120% of the overall variation (Table 3).
Confirmatory Factor Analysis
To ascertain the FHAI’s criteria-dependent validity, the NuPDQ was used as the criterion. According to the results of the correlation analysis, there was a statistically significant positive association between the FHAI total score and the NuPDQ overall score as well as its subscale total scores (p=0.01; Table 4).
According to the results of CFA performed on Sample II, the values of c2/sd (cmin/df) =3.148, CFI =0.907, SRMR =0.000, RMSEA =0.089, and p value =0.000 were determined for the FHAI (Table 4; Figure 2). It was found that there was a satisfactory fit between the measurement model and the data after looking at the numbers in the fit indicators (Table 5).
Reliability
Cronbach’s Alpha Reliability Coefficient
The FHAI was found to have an internal consistency coefficient of Cronbach’s alpha of 0.85 (p<0.001; Table 3).
Item-total Score Correlation Coefficient
It was discovered that each of these items had a correlation coefficient with the overall score derived from the inventory items that ranged from r=0.34-0.59 and that this connection was statistically significant (p<0.001; Table 3).
Test-retest Mean Scores of the FHAI
A statistically significant positive correlation between the total mean scores of the inventory was discovered when the first and second measurement correlation findings of the FHAI administered with a three-week interval were assessed (r=0.568, p=0.001).
Discussion
Chronic stress is one of the most common modifiers of fetal and postnatal development with lifelong effects on health (37). There is no scale to determine fetal anxiety stress in Turkey. The FHAI was modified for Turkish use in the current study, and the instrument’s reliability and validity were assessed. With a sample of Turkish pregnant women, the FHAI showed overall high validity and reliability. The CFA in this study verified that the single-factor construct had good fit indices and its reliability values were in an acceptable range. The EFA in this study suggested that a pregnant mother’s worry for the health of her growing fetus might be analyzed as a single-factor construct. It was found that the factor loads of the FHAI in the Turkish adaption research were distributed similarly to how they were in the original form. The data of the CFA fit index, which was used to check whether the items were compatible with the data in the original FHAI (21), was not used, but the results of the analysis we used in the Turkish version were in the right range of values, ensuring the construct validity of the inventory (31,36,38).
In the present study, factor loads of the FHAI varied between 0.41-0.69 according to the results of EFA. Similarly, factor loading values of the FHAI were found to be between 0.51 and 0.77 in the original inventory (22). It was discovered that the distribution of the items in the Turkish inventory, which was made up of the initial 14 items, was the same. As a result of the EFA applied in the original Reiser and Wright’s (21) inventory, it was possible to divide the inventory into two subscales titled Disease Probability and Body Attention, but in a two-factor structure, item 1 (time worrying about your baby’s health) and item 7 (ability to take the mind off of thoughts of baby’s health) indicated that the factors were not distributed appropriately. Considering this inconsistency, it was suggested that items 1 and 7 might not be loaded on individual subscales as expected, and they could be used as a unitary inventory as well as subscales of the FHAI (21). In the present study, it was determined that Reiser and Wright’s single-factor model fitted the Turkish FHAI data more appropriately. Additionally, it was discovered that the KMO, Barlett’s test of sphericity, and explained variance values in this investigation were comparable to those in the initial study. The data in the current investigation showed an adequate distribution for factor analysis, according to the results of KMO and Barlett’s test of sphericity (39). The overall variance of the inventory can be deemed sufficient when 30% or more of the variation rate described in scale adaptation experiments is regarded sufficient (28).
Strong relationships between prenatal distress (NuPDQ total score, including the subscale of concerns regarding the baby) and the FHAI total score were found, indicating high concurrent validity (40). Additionally, Reiser and Wright (21) showed relationships between the FHAI and measures of anxiety related to pregnancy, anxiety related to maternal health, general anxiety, anxiety sensitivity, and uncertainty intolerance. Numerous studies in the literature demonstrate the detrimental short- and long-term consequences of anxiety on the mother and unborn child during pregnancy (41-43). The significance of the function that fetal health anxiety can play in mother’s and child’s health has been shown when taking into account the association between prenatal distress and fetal health anxiety in the current investigation.
In the current study, the FHAI showed good internal consistency. The FHAI, developed by Reiser and Wright (21), was also reported to have a high level of reliability when Cronbach’s alpha value was examined. In addition, the item-total score correlation coefficients in the current study were found to be similar to those in Reiser and Wright’s (21) study, and it was determined that the coefficients of all items were above the acceptable value (r≥0.20 for all items in the inventory) for item selection (33). To prove that the FHAI was invariant over time, the inventory was re-administered to 30 pregnant women three weeks after the first application. The high correlation value between the first and second application scores indicates that the inventory gives consistent results and is invariant over time.
Given these findings, it can be concluded that the FHAI, of which Turkish validity and reliability research we have conducted, and the original of this inventory are in agreement, making it a valid and reliable instrument for assessing pregnant women’s fetal health anxiety.
Study Limitations
The present study had certain shortcomings. First off, pregnant women from other institutions were not included in this study since it only included pregnant women who were registered at one public hospital. The findings of this study might not apply to all pregnant women because it only included pregnant women who were admitted to the hospital. Another drawback was that the FHAI validity and reliability analyses were conducted without taking gestation and trimester into account. Future research may advocate doing the validity and reliability tests of the inventory independently based on gestational age and trimester. Regarding the measurement, web-based questionnaires were applied to pregnant women, and the reports were in the form of self-reports. This variable was susceptible to bias.
Conclusion
The FHAI is a potential tool for detecting fetal health anxiety during pregnancy, as well as for application in research and clinical practice, according to the study’s findings. Due to the harmful effects that may occur on maternal and fetal health of those who experience fetal health anxiety intensely, healthcare professionals can make a better analysis of the expectant mothers’ anxiety levels in order to be aware of the anxiety levels experienced by women during pregnancy, and they can enable the mother and fetus to have a better pregnancy period. This may also assist in developing specific initiatives to meet the needs of individual care. The FHAI can be applied as a rapid and accurate pre-screening tool to assess fetal health anxiety levels in the clinics and studies.