Effect of Educational Intervention Based on PRECEDE Model on Improving Menstrual Health Behaviors Among Female Students

authors:

avatar Mohammadali Rezaei 1 , avatar Laleh Solaimani ORCID 2 , avatar Neda Mohammadinia 1 , *

Department of Nursing, School of Nursing & Midwifery, Bam University of Medical Sciences, Bam, Iran
Department of Community Health Nursing, School of Nursing and Midwifery, Bam University of Medical Sciences, Bam, Iran

how to cite: Rezaei M, Solaimani L, Mohammadinia N. Effect of Educational Intervention Based on PRECEDE Model on Improving Menstrual Health Behaviors Among Female Students. Shiraz E-Med J. 2022;23(12):e123461. https://doi.org/10.5812/semj-123461.

Abstract

Background:

Poor menstrual health is a risk factor for genital infections and infertility. It may affect family health and the community’s future.

Objectives:

This study aimed to determine the effect of educational intervention based on the PRECEDE model on improving menstrual health behaviors among female students.

Methods:

This quasi-experimental study was conducted on 130 first-grade high school female students and their mothers randomly selected from six schools and divided into intervention and control groups. Three two-hour educational sessions were conducted for the students and their mothers in the intervention group based on the pretest results. Data were collected immediately and three months after the educational sessions using a questionnaire with confirmed validity and reliability. Data were analyzed using SPSS v. 21 with repeated-measures ANOVA and t-test. The significance level was considered at 0.05.

Results:

The educational intervention based on the PRECEDE model increased the mean scores of predisposing, enabling, and reinforcing factors in the intervention group and finally increased the mean score of health behavior immediately (24.1 ± 4.62, P < 0.001) and three months after the intervention (22.84 ± 4.04, P = 0.02) compared to before intervention (19.68 ± 5.48).

Conclusions:

The results showed a change in factors affecting behavior (predisposing, enabling, and reinforcing factors). The PRECEDE model changed the menstrual health behaviors among the students.

1. Background

Adolescence is a critical period in which puberty takes place, which is the time of gaining the ability to fertility (1). The onset of menstruation is the most prominent benchmark and significant event in girls’ lives. Menstruation is the most critical developmental phenomenon in girls’ reproductive and fertility systems (2). In addition, during menstruation, girls face numerous challenges. An average woman experiences 400 menstrual periods from menarche to menopause (3). According to the World Health Organization (WHO), mental and physical symptoms in the menstrual period and dysmenorrhea have a prevalence rate of 1.7% to 97% (4). In many cases, it leads to short-term absences from school, disruption of daily activities (5), reduced student participation in social and educational activities, and eventually, leaving school (6).

In addition, numerous studies have shown that many girls were uneducated about menstruation before menarche (7), and their knowledge of puberty, menstrual physiology, and hygiene was superficial and inadequate (1). Their menstrual health behaviors were often wrong and associated with false beliefs (8). Also, in Iran (9-11) and some neighboring countries, different behaviors, beliefs, and dietary, social, and religious limitations exist in this respect (12-14).

Therefore, given the lack of education, misinformation, shame, and avoidance of discussing genital health in this period, which precedes a successful transition to adulthood and fertility, as well as poor menstrual health, it is a risk factor for genital infections and infertility and may affect family and community health (15). Menstrual health is integral to realizing gender equality and human right and achieving sustainable development goals (16).

In today’s world, adolescent health is considered an important issue. It has been much emphasized, especially since the International Population and Development Conference in 1994. The WHO particularly addresses the educational needs of adolescent girls (17). One of the main concerns of health authorities is proper health education because the value of health education programs depends on their effectiveness, which is mainly dependent on the correct use of the theories and models (18). The present study used the PRECEDE model for health education and interventions, designed by Marshall Crowther and Lawrence Green in 1980 (19).

The PRECEDE model is one of the most applicable and appropriate models for behavior change that explores the possible outcome of an educational program (20). This model is easy to use for different subjects and populations. Also, in this model, behavioral change depends not only on one’s functioning but also on the environment (19). This model provides a framework consisting of factors influencing behavior, such as predisposing factors (knowledge and attitude), reinforcing factors, (encouragement by family, peers, and teachers), and enabling factors (educational resources, classes, facilities, equipment, etc.) (21), which are formed by the PRECEDE form. PRECEDE stands for Predisposing, Reinforcing, and Enabling Constructs in Educational Diagnosis and Evaluation of target population and community (20).

2. Objectives

This study aimed to determine the effect of educational intervention based on the PRECEDE model on improving menstrual health behaviors of 13-16-year-old female students in Bam city.

3. Methods

This is a quasi-experimental study with a pretest-posttest design. The study population consisted of 13 - 16-year-old high school students and their mothers in Bam city in the east of Iran in 2019. Inclusion criteria for the students consisted of willingness to participate in the study, being present in educational interventions, completing the questionnaires, and having at least three menstrual cycles. The study sample was selected from eligible students according to the formula for comparing the means of two independent groups. This formula used a statistical power of 80%, a confidence coefficient of 0.95, a between-group difference of 1.92, and a practice score standard deviation of 6.27 (21). The sample size was 56 in each of the intervention and control groups. By considering at least 10% dropout, the required sample size was 60 individuals per group. First, four out of 25 high schools were selected via simple random sampling. Then, two schools for intervention and two for control were randomly selected. Finally, 30 students and their mothers were selected from each school (10 students from each grade). These students were selected based on the random number table and school lists.

A two-part self-administered questionnaire was designed. The first part included demographic characteristics, and the second part concerned educational-environmental evaluations based on the precede model structures. Items in each section were individually scored and designed so that predisposing factors included 15 knowledge items with the scores 0 - 2 (2 = correct, 1 = unknown, 0 = incorrect) and 20 attitude items with the Likert scale (4 = completely agree, 3 = agree, 2 = indifferent, 1 = disagree, 0 = completely disagree). Moreover, the enabling factors in home and school had 24 items with the scores 0 - 1 (1 = yes and 0 = no) and 15 behavior items with the scores 0 - 2 (2 = always, 1 = some times, 0 = never). In this study, reinforcing factors included a collection of mothers’ knowledge, attitude, and practice regarding menstrual health behaviors. All negative questions were scored inversely.

A panel of 14 experts in gynecology, reproductive health, and health education assessed the questionnaire’s validity. The CVR was determined using the Lawshe table, the questions with CVR > 0.59, and the questions with CVI > 0.79 were accepted, and the questions with CVI > 0.79 were accepted.

Also, the reliability of the questionnaire was confirmed through the implementation among some of the target groups (20 students and 15 mothers of the same students) and calculating Cronbach’s alpha coefficients (α = 0.85 for predisposing factors, α = 0.84 for enabling factors, and α = 0.80 for behavior section).

First, each student and her mother filled out the pretest questionnaire separately. After analyzing the pretest scores and an extensive review of relevant resources, educational materials were formulated and confirmed based on the opinions of four experts in health education. The education was conducted in three two-hour sessions using lecture, face-to-face, discussion, and question/answer methods for students and mothers in the intervention group. The educational sessions were about adolescence, puberty, t the menstrual cycle, abnormal signs and common problems associated with menstruation, menstrual health, exercise, nutrition, mobility, and pain control in menstruation. Immediately after the last session and three months later, the posttest questionnaires were distributed among the research subjects in the absence of the researcher. Finally, the data obtained before and after the intervention were analyzed in separate groups using SPSS v.21 software with repeated-measures ANOVA and t test.

4. Results

The data of 56 students and 55 mothers in the control group and 58 students and 57 mothers in the intervention group were analyzed. The mean age of participants was 14.8 ± 1.07 years in the intervention group and 14.21 ± 1.11 years in the control group, but the t-test and χ2 test showed that demographic characteristics were not significantly different between the two groups before the education, so the control and intervention groups were homogeneous (Table 1). The t test showed that the mean scores of the PRECEDE model constructs were not significantly different between the two groups before the educational intervention (P > 0.05). The scores of all constructs increased significantly after the educational intervention compared to before and not only remained stable but also increased during the follow-up period (P < 0.05). The repeated-measures ANOVA showed that the mean scores of students on all structures had a significant difference between the pretest and posttest stages in the intervention group.

Table 1.

Demographic Characteristics in Two Groups Before Intervention

VariableIntervention GroupControl GroupP-Value
Father’s occupation status0.22
Employed43 (71.6)37 (61.7)
Unemployed15 (25)19 (31.6)
Mother’s occupation Status0.31
Employed19 (31.6)21 (35)
Unemployed39 (65)35 (58.4)
Father’s education0.33
Illiterate3 (5)4 (6.7)
Primary school5 (8.3)2 (3.3)
Secondary school8 (13.3)11 (18.3)
High school17 (28.3)21 (35)
University25 (41.6)18 (30)
Mother’s education0.27
Illiterate4 (6.7)2 (3.3)
Primary school11 (18.3)9 (15)
Secondary school8 (13.3)13 (21.7)
High school22 (36.6)17 (28.3)
University13 (21.7)15 (25)

In addition, the t test showed that the scores of all constructs immediately and three months after the intervention (except for knowledge in the follow-up stage) were significantly higher in the intervention group than in the control group (P < 0.05). The mean score of menstrual health behavior was significantly higher in the intervention group than in the control group, immediately (P < 0.001), and three months after intervention (P = 0.02) (Table 2).

Table 2.

PRECEDE Model Constructs Before and After Educational Intervention in Two Groups

ConstructsIntervention GroupControl GroupP-Value a
Knowledge
Before intervention21.11 ± 6.1221.82 ± 5.160.19
Immediately after intervention24.1 ± 4.2322.84 ± 4.180.04
Three months after intervention23.6 ± 3.1522.81 ± 3.260.286
P-value bP < 0.0010.37
Attitude
Before intervention25.61 ± 4.6125.48 ± 3.120.857
Immediately after intervention27.11 ± 4.2525.68 ± 3.340.043
Three months after intervention27.86 ± 4.1925.71 ± 3.130.002
P-value bP < 0.0010.32
Enabling Factors
Before intervention32.45 ± 6.4833.14 ± 3.580.22
Three months after intervention39.61 ± 5.7332.85 ± 3.910.03
P-value a< 0.0010.22
Reinforcing Factors
Before intervention61.02 ± 5.160.91 ± 6.10.21
Immediately after intervention72.11 ± 3.1861.37 ± 6.8< 0.001
Three months after intervention70.12 ± 3.2761.07 ± 6.14< 0.001
P-value b< 0.0010.17
Behavior
Before intervention19.68 ± 5.4819.21 ± 4.180.51
Immediately after intervention24.1 ± 4.6220.4 ± 5.34< 0.001
Three months after intervention22.84 ± 4.0420.19 ± 4.450.02
P-value b< 0.0010.41

5. Discussion

The results showed that education based on the PRECEDE model effectively increased the mean scores of variables and, ultimately health behavior of the intervention group. The high scores of the variables three months after education in the intervention group affirmed the reliability of their health behavior. Although no studies were found using the PRECEDE model to enhance menstrual health behavior, several studies confirmed the effect of the model on behavior change in other subjects (18-20, 22).

The PRECEDE model includes the predisposing (knowledge and attitude), reinforcing, and enabling factors in the evaluation/diagnosis of education/environment (23). In the present study, these factors were associated with the necessary information on menstrual health care concerning personal hygiene, nutrition, activity, menstrual pain relief, and students’ attitude about menstruation and caring methods. In this study, the mean scores of predisposing factors (knowledge and attitude) were significantly increased in the intervention group after the education, which is in line with the results of similar studies performed using the structures of the PRECEDE model to correct physical activity (18), improve self-care behaviors (24), self-manage type 2 diabetes mellitus (25).

The results showed many students had high awareness of underwear hygiene (84%) and how to dispose of sanitary napkins (89%), while the awareness of bathing (91.2%) and mobility during menstruation was poor (63%) in most of them. The majority of the participants (97%) believed that menstrual hygiene prevented infection, while more than half (57.7%) considered that consuming cold-natured, flatulent, and sour foods during menstruation caused menstrual pain. Also, 59.3% thought that exercises (heavy exercise, running, jumping, and jumping rope) caused prolapsed uterus. Their knowledge and attitude were significantly improved after the educational intervention. In similar studies, subjects also pointed out contaminated menstrual blood and bleeding women (12) and alluded to the limitations of exercise and intense activity to staying at home (26), restriction house chores (12, 27)and forbidding some foods (fried, sweet, sour, and spicy foods) (26, 28). As for bathing during menstruation, studies have reported mixed results, including believing in daily bathing during menstruation (26) and bathing prohibition during menstruation (26, 28). These differences may be attributed to their cultural attitude, beliefs, and level of knowledge. Hence, school education classes can provide context to solve problems, answer questions, and improve awareness and attitude, which are predisposing factors for initiating and sustaining health behavior.

The second area of the model includes reinforcing factors, i.e., factors that encourage one to change behavior (23). In this study, reinforcing factors included the mothers’ knowledge, attitude, and practice regarding menstrual health behaviors. This study showed a significant difference between the mean scores of reinforcing factors in the intervention group before and after the intervention compared to the control group, which is in line with other research (18, 23, 29).

The results of the present study (79.26%) and other similar studies confirm that the mother is the most important source of information on menstrual and puberty problems among adolescents (20, 30), followed by peers (43.8%) and teachers (42.9%). Also, most of them (79.26%) had several sources of information in this area. The close relationship between mother and daughter can be a good reason for this result. Hence, such issues as enhancing the relationship between mother and daughter, addressing barriers such as shyness in expressing puberty and menstruation problems, or ignorance of the relevant health consequences by mothers should be included in educational programs for mothers. Therefore, mothers should be taught how to make friendly relations with their daughters to attract the trust of their girls, how to pay attention to their physical, psychological, mood, and behavioral changes during adolescence, and how to prevent negative consequences and utilizing the strengths of this period, as the best strategy to increase health knowledge and health behavior in girls is educating families, especially mothers (31). The mean scores of enabling factors (such as access to sanitary and hygiene services and materials, menstrual counseling, training sessions, and appropriate educational resources) were significantly increased three months after the educational intervention because enabling factors cannot change immediately after education, and they need time, which was in line with similar studies (18, 23-27, 32)Of course, educating mothers could effectively increase the scores of enabling factors.

In the present study, with the significant increase in the mean scores of the constructs affecting behavior, the students’ health behaviors also increased significantly after the educational intervention and in the follow-up period compared to the control group, in line with similar studies (18, 24, 25). The mean behavior score immediately after the educational intervention was higher than the mean behavior score three months after the intervention, confirming the need for continuity and repetition of training sessions. Thus, using specific models, such as PRECEDE, in educational programs can help explain what needs to be understood, such as recognizing environmental and educational factors.

This study was applied to the students and their mothers in Bam city, so the results may not be generalized to other adolescents. The menstruation subject sometimes has embarrassment or shame, influencing the self-report results. Being acquainted with the researcher and the study objectives, the nature of the problem, and no coercion to participate in the study were employed to overcome the study’s limitations.

5.1. Conclusion

Designing and implementing educational interventions based on the PRECEDE model can modify health behaviors during menstruation. Our study confirmed that not only the person’s but also the knowledge, attitude, and behavior of the family (reinforcing factors) and the resources available (enabling factors), such as menstrual hygiene materials, sanitary and hygiene services, and safety and purified water, can affect menstrual health behavior and should pay attention to in health programs. Also, the results of the follow-up period showed the long-term effectiveness and reliability of the behavior with this model.

Acknowledgements

References

  • 1.

    Afshary P, Pazhohideh SZ, Yazdi Zadeh H, Mohammadi S, Tabesh H. [Survey educational needs of 11-14 years old Girls about Sexual health]. J Holist Nurs Midwifery. 2016;26(1):1-9. Persian.

  • 2.

    Brêtas JRDS, Tadini AC, Freitas MJDD, Goellner MB. Significado da menarca segundo adolescentes. Acta Paul Enferm. 2012;25(2):249-55. https://doi.org/10.1590/s0103-21002012000200015.

  • 3.

    Siniša F. Menstrual pain. Journal of Gynecological Research and Obstetrics. 2019;5(2):31-3. https://doi.org/10.17352/jgro.000067.

  • 4.

    Mohamadirizi S, Kordi M, Shakeri MT. The relationship between perimenstrual symptoms and menstrual attitude in high-school girls in Mashhad city in years 2011-2012. Iran J Obstet Gynecol Infertil. 2012;15(8):25-31.

  • 5.

    Mohammadinia N, Rezaei M, Salehian T, Dashipoor A. [Comparing the effect of Anethum gravolens with mefenamic acid consumption on treatment of primary dysmenorrhea]. J Shahrekord Univ Med Sci. 2013;15(5):57-64. Persian.

  • 6.

    Boosey R, Prestwich G, Deave T. Menstrual hygiene management amongst schoolgirls in the Rukungiri district of Uganda and the impact on their education: a cross-sectional study. Pan Afr Med J. 2014;19:253. [PubMed ID: 25852796]. [PubMed Central ID: PMC4382073]. https://doi.org/10.11604/pamj.2014.19.253.5313.

  • 7.

    Sharma S, Mehra D, Brusselaers N, Mehra S. Menstrual Hygiene Preparedness Among Schools in India: A Systematic Review and Meta-Analysis of System-and Policy-Level Actions. Int J Environ Res Public Health. 2020;17(2). [PubMed ID: 31963862]. [PubMed Central ID: PMC7013590]. https://doi.org/10.3390/ijerph17020647.

  • 8.

    Amatya P, Ghimire S, Callahan KE, Baral BK, Poudel KC. Practice and lived experience of menstrual exiles (Chhaupadi) among adolescent girls in far-western Nepal. PLoS One. 2018;13(12). e0208260. [PubMed ID: 30532183]. [PubMed Central ID: PMC6287853]. https://doi.org/10.1371/journal.pone.0208260.

  • 9.

    Vakilian K, Davod Abadi M, Gholami R. [Exersise corelation with pattern dysmenoreha in femal student of Arak medical universityin 1389-90]. J Complement Med Res. 2015;5(3):1214-23. Persian.

  • 10.

    Davaneghi S, Tarighat Esfanjani A, Dahri M. [Association of nutritional factors and physical activity with severity of primary dysmenorrheal pain]. J Health Care. 2016;18(2):93-100. Persian.

  • 11.

    Mobin E, Mirzaei M, Karimi M. [Investigating knowledge, attitude, and health behaviour of high school female students concerning menstrual hygiene]. Tolooebehdasht. 2014;12(4):70-8. Persian.

  • 12.

    Srivastava S, Chandra M. Study on the knowledge of school girls regarding menstrual and reproductive health and their perceptions about family life education program. Int J Reprod Contracept Obstet Gynecol. 2017;6(2):688. https://doi.org/10.18203/2320-1770.ijrcog20170406.

  • 13.

    Fehintola FO, Fehintola AO, Aremu AO, Idowu A, Ogunlaja OA, Ogunlaja IP. Assessment of knowledge, attitude and practice about menstruation and menstrual hygiene among secondary high school girls in Ogbomoso, Oyo state, Nigeria. Int J Reprod Contracept Obstet Gynecol. 2017;6(5):1726. https://doi.org/10.18203/2320-1770.ijrcog20171932.

  • 14.

    Tamiru S, Sahin M, Mamo K, Acidria P, Mushi R, Ali CS, et al. Towards a sustainable solution for school menstrual hygiene management: cases of Ethiopia, Uganda, South-Sudan, Tanzania, and Zimbabwe. Waterlines. 2015;34(1):92-102. https://doi.org/10.3362/1756-3488.2015.009.

  • 15.

    Eslamimehr F, Ramezankhani A, Khodakarim S, Rakhshani F. [Assessment of puberty health behaviour and it’s affecting factors in female secondary school students in Khamir city during 2015-2016]. Journal of Health in the Field. 2017;4(2):11-8. Persian.

  • 16.

    Hennegan J, Winkler IT, Bobel C, Keiser D, Hampton J, Larsson G, et al. Menstrual health: a definition for policy, practice, and research. Sex Reprod Health Matters. 2021;29(1):1911618. [PubMed ID: 33910492]. [PubMed Central ID: PMC8098749]. https://doi.org/10.1080/26410397.2021.1911618.

  • 17.

    Coast E, Lattof SR, Strong J. Puberty and menstruation knowledge among young adolescents in low- and middle-income countries: a scoping review. Int J Public Health. 2019;64(2):293-304. [PubMed ID: 30740629]. [PubMed Central ID: PMC6439145]. https://doi.org/10.1007/s00038-019-01209-0.

  • 18.

    Khaleghi Mahani H, Ahmadi Tabatabaei SV, Mirzai M, Poursharifei A. [Breast Self–examination Predictors based on Precede Model: a study on female active health volunteers in Zarand City]. Journal of Health Based Research. 2017;3(1):71-85. Persian.

  • 19.

    Tabasi Nejad N, Mohseni M, Khanjani N, Ahmadi Tabatabaei SV. Application of the Precede-Proceed Model in Promoting Physical Activity for Prevention of Osteoporosis among Women. Health Education and Health Promotion. 2018;6(3):103-8. https://doi.org/10.29252/hehp.6.3.103.

  • 20.

    Mosavi F, Aliakbari F, Rabiei L. Effect of education based on "PRECEDE" model on self-care behavior in hemodialysis patients. J Educ Health Promot. 2020;9:69. [PubMed ID: 32490004]. [PubMed Central ID: PMC7255572]. https://doi.org/10.4103/jehp.jehp_534_19.

  • 21.

    Fadaei E, Borhani M, Hoseini ZS, Mehri A, Tatari M. Effect of Health Educational Intervention based on the Educational Phase of the PRECEDE-PROCEED Model on the Promotion of Preventive Behaviors of Brucellosis in the Villagers of Minoodasht, Iran. J Educ Community Health. 2021;8(3):203-8. https://doi.org/10.52547/jech.8.3.203.

  • 22.

    Barasheh N, Shakerinejad G, Nouhjah S, Haghighizadeh MH. The effect of educational program based on the precede-proceed model on improving self-care behaviors in a semi-urban population with type 2 diabetes referred to health centers of Bavi, Iran. Diabetes Metab Syndr. 2017;11 Suppl 2:S759-65. [PubMed ID: 28669595]. https://doi.org/10.1016/j.dsx.2017.05.012.

  • 23.

    Khani Jeihooni A, Heidari MS, Afzali Harsini P, Azizinia S. Application of PRECEDE model in education of nutrition and physical activities in obesity and overweight female high school students. Obes Med. 2019;14:100092. https://doi.org/10.1016/j.obmed.2019.100092.

  • 24.

    Shahnazi H, Tabar Isfahani M, Azarbin S, Hassanzadeh A, Charkazi A, Moodi M. [Impact of education based on precede model on knowledge, attitude and behavior of grade two guidance school girls regarding Iron Deficiency Anemia (IDA) in Isfahan, Iran]. J Health Res. 2012;18(5):773-81. Persian.

  • 25.

    Nejhaddadgar N, Darabi F, Rohban A, Solhi M, Kheire M. The effectiveness of self-management program for people with type 2 diabetes mellitus based on PRECEDE-PROCEDE model. Diabetes Metab Syndr. 2019;13(1):440-3. [PubMed ID: 30641740]. https://doi.org/10.1016/j.dsx.2018.08.016.

  • 26.

    Sadiq MA, Salih AA. Knowledge and practice of adolescent females about menstruation in Baghdad. J Gen Pract. 2013;2(1):1000138. https://doi.org/10.4172/2329-9126.1000138.

  • 27.

    El-Lassy RB, Madian AAE. Impact of health educational program on menstrual beliefs and practices of adolescent Egyptian girls at secondary technical nursing school. Life Sci. 2013;10(2):1-5.

  • 28.

    Morowatisharifabad MA, Vaezi A, Bokaie M, Askarishahi M, Mohammadinia N. Cultural beliefs on menstrual health in Bam city: a qualitative study. Int J Pediatr. 2018;6(12):8765-78.

  • 29.

    Sanjari S, Mohammadi Soleimani MR. [The impact of education on the prevention of vaginal infection based on the PRECEDE-PROCEED model]. J Rafsanjan Univ Med Sci. 2017;16(4):338-52. Persian.

  • 30.

    Abedi R, Tol A, Majlessi F, Mahmoodi Majdabadi M. [Determining the effect of an educational intervention based on constructs of Health Belief Model on Promotion of Physical Puberty Health among student girls]. Razi Journal of Medical Sciences. 2015;22(137):84-94. Persian.

  • 31.

    Mostofi N, Garmaroudi G, Shamshiri A, Shakibazadeh E. Effect of group education on knowledge, attitude and practice of mothers of adolescent girls about sex education. Journal of School of Public Health and Institute of Public Health Research. 2016;14(3):45-59.

  • 32.

    Solhi M, Shabani Hamedan M, Salehi M. A PRECEDE-PROCEED based educational intervention in quality of life of women-headed households in Iran. Med J Islam Repub Iran. 2016;30:417. [PubMed ID: 28210582]. [PubMed Central ID: PMC5307607].