JMIR mHealth and uHealth
Mobile and tablet apps, ubiquitous and pervasive computing, wearable computing and domotics for health.
JMIR mhealth and uhealth (mobile and ubiquitous health) (JMU, ISSN 2291-5222) is a new spin-off journal of JMIR, the leading eHealth journal (Impact Factor 2014: 3.4). JMIR mHealth and uHealth has a projected impact factor (2015) of about 2.03. The journal focusses on health and biomedical applications in mobile and tablet computing, pervasive and ubiquitous computing, wearable computing and domotics.
JMIR mHealth and uHealth publishes even faster and has a broader scope with including papers which are more technical or more formative/developmental than what would be published in the Journal of Medical Internet Research.
In addition to peer-reviewing paper submissions by researchers, JMIR mHealth and uHealth offers peer-review of medical apps itself (developers can submit an app for peer-review here).
JMIR mHealth and uHealth features a rapid and thorough peer-review process, professional copyediting, professional production of PDF, XHTML, and XML proofs.
JMIR mHealth and uHealth adheres to the same quality standards as JMIR and all articles published here are also cross-listed in the Table of Contents of JMIR, the worlds' leading medical journal in health sciences / health services research and health informatics.
May 6, 2016
May 5, 2016
Apr 27, 2016
Apr 25, 2016
Apr 22, 2016
Apr 20, 2016
Apr 19, 2016
Apr 15, 2016
Apr 15, 2016
Apr 13, 2016
Apr 11, 2016
Apr 7, 2016
Citing this Article
Right click to copy or hit: ctrl+c (cmd+c on mac)
Latest Submissions Open for Peer-Review:View All Open Peer Review Articles
Engaging Gatekeeper-stakeholders in Development of a Mobile Health Intervention to Improve Medication Adherence among African American and Pacific Islander Elderly Patients with Hypertension
Date Submitted: Apr 25, 2016
Open Peer Review Period: Apr 29, 2016 - Jun 24, 2016
Background: Approximately 70 million people in the US have hypertension. While antihypertensive therapy can reduce the morbidity and mortality associated with hypertension, often patients do not take...
Background: Approximately 70 million people in the US have hypertension. While antihypertensive therapy can reduce the morbidity and mortality associated with hypertension, often patients do not take their medication as prescribed. Objective: The goal of this study was to better understand issues affecting the acceptability and usability of mobile health technology (mHealth) to improve medication adherence for elderly African American (AA) and Native Hawaiian and Pacific Islander (NHPI) patients with hypertension. Methods: In-depth interviews were conducted with 20 Gatekeeper-stakeholders using targeted open-ended questions. Interviews were de-identified, transcribed, organized and coded manually by two independent coders. Analysis of patient interviews used largely a deductive approach because the targeted open-ended interview questions were designed to explore issues specific to the design and acceptability of a mHealth intervention for seniors. Results: A number of similar themes regarding elements of a successful intervention emerged from our two groups of AA and NHPI Gatekeeper-stakeholders. First was the need to teach participants both about the importance of adherence to antihypertensive medications; second, the use of smart/cell phones for messaging and patients need to be able to access ongoing technical support; third, messaging needs to be short and simple, but personalized, and to come from someone the participant trusts and with whom they have a connection. There were some differences between groups. For instance, there was a strong sentiment among AA that the church be involved and that the intervention begin with group workshops, while NHPI seemed to believe that the teaching could occur on a one-to-one basis with the health care provider. Conclusions: Information from our Gatekeeper-stakeholder (key informant) interviews suggests that the design of the mHealth intervention to improve adherence to antihypertensives among the elderly could be very similar between AAs and NHPIs. The main difference might be in the way in which the program is initiated (possibly through church-based workshops for AA and by individual providers for NHPIs). Another difference might be who sends the messages with AA wanting someone outside the health care system, but NHPI preferring a provider.
Investigating the Perceptions of Care Coordinators on Using Behavior Theory-Based Mobile Health Technology with Medicaid Populations: A Grounded Theory Study
Date Submitted: Apr 20, 2016
Open Peer Review Period: Apr 22, 2016 - Jun 17, 2016
Background: Medicaid populations are less engaged in their healthcare than the rest of the population, translating to worse health outcomes and increased healthcare costs. Since theory-based mobile he...
Background: Medicaid populations are less engaged in their healthcare than the rest of the population, translating to worse health outcomes and increased healthcare costs. Since theory-based mobile health (mHealth) interventions have been shown to increase patient engagement, mobile phones may be an optimal strategy to reach this population. There is a deep disconnect between developers of mHealth technology and health behavior researchers, so there is a lack of data on what components of theory-based mHealth increase patient engagement. Objective: This study aims to address this gap between academia and practice by conducting research using the health behavior-theory based patient-provider text-messaging platform, Sense Health, which integrates Transtheoretical Model and Stages of Change (TTM), Social Cognitive Theory (SCT), Supportive Accountability, and Motivational Interviewing. Methods: Interviews based in grounded-theory methodology were conducted with 10 care managers to triangulate the findings of internal user activity data and to further understand perceptions of the relationship between mHealth and patient engagement. Results: The interviews with care managers yield a grounded theory model including four intertwined relationships revolving around patient engagement: between Sense Health, client-care manager relationships, and communication; Sense Health, literacy, and access to care; support, Sense Health, and communication; and Sense Health, patient accountability, and patient motivation. Conclusions: Sense Health features tied to health behavior theory appear to be effective in improving patient engagement. Two-way communication (Supportive Accountability), trusted relationships (Supportive Accountability, SCT), personalized messages (TTM), and patient input (TTM, SCT, Motivational Interviewing) appeared as the most relevant components in achieving desired outcomes. Additionally, reminder messages were noted as especially useful in making Medicaid patients accountable, and in turn engaging them in their health and healthcare. These findings expose how this theory-centered platform drives engagement, allowing Sense Health, and future mHealth interventions that aim to improvement patient engagement in Medicaid populations, to improve their technology. Clinical Trial: Columbia University Medical Center Institutional Review Board (IRB-AAAQ5254)
SMS-Based Intervention Targeting Alcohol Consumption Among University Students: Findings from a Formative Development Study
Date Submitted: Apr 12, 2016
Open Peer Review Period: Apr 12, 2016 - Jun 7, 2016
Background: Drinking of alcohol among university students is a global phenomenon with heavy episodic drinking being accepted despite several potential negative consequences. Half of all young adults i...
Background: Drinking of alcohol among university students is a global phenomenon with heavy episodic drinking being accepted despite several potential negative consequences. Half of all young adults in Sweden attend university making the health and well-being of this group a public health concern. There is emerging evidence that text messaging (SMS) interventions are effective to promote behaviour change among students. However, it is still unclear how effectiveness can be optimized through intervention design or how user interest and adherence can be maximised. Objective: To develop an SMS-based intervention targeting alcohol drinking among university students using formative research. Methods: A formative research design was used including an iterative revision process based on input from end users and experts. Data were collected via focus groups (n=7) with students and a panel evaluation involving students (n=15) and experts (n=5). Student participants were recruited from five universities in Sweden. A semi-structured interview guide was used in the focus groups and included questions on alcohol culture, message content and intervention format. The panel evaluation asked participants to rate to what degree preliminary messages were understandable, usable and had a good tone on a scale from 1 to 4 (1 = very low degree; 4 = very high degree). Participants could also write their own comments for each message. Qualitative data were analysed using qualitative descriptive analysis. Quantitative data were analysed using descriptive statistics. The SMS messages and the intervention format were revised continuously, in parallel with data collection. A behaviour change technique analysis was conducted on the final version of the program. Results: The focus group data showed that, overall, students were positive towards the SMS intervention. Messages that were neutral, motivated, clear and tangible engaged students. Students expressed that they preferred short, concise messages and confirmed that a 6-week intervention was an appropriate duration. However, there was limited consensus regarding SMS frequency, personalization of messages and timing. Overall, messages scored high on understanding (3.86, SD 0.43), usability (3.70, SD 0.61) and tone (3.78, SD 0.53). Participants added comments to 67 of 70 messages, including suggestions for change in wording, order of messages, and feedback on why a message was unclear or needed major revision. Comments also included positive feedback that confirmed the value of the messages. Twenty-three behaviour change techniques, aimed at, for example, addressing self-regulatory skills, were identified in the final program. Conclusions: The formative research design was valuable and resulted in significant changes to the intervention. All the original SMS messages were changed and new messages were added. The findings showed that, overall, students were positive towards receiving support through SMS and that neutral, motivated, clear and tangible messages promoted engagement. However, limited consensus was found on the timing, frequency and tailoring of messages.
A mobile phone app for dietary intake assessment (e-EPIDEMIOLOGY): comparison with a food frequency questionnaire
Date Submitted: Mar 22, 2016
Open Peer Review Period: Mar 22, 2016 - May 17, 2016
Background: Background: There is a great necessity for new methods of evaluation of dietary intake that overcome the limitations of traditional methods such as food frequency questionnaires (FFQ). O...
Background: Background: There is a great necessity for new methods of evaluation of dietary intake that overcome the limitations of traditional methods such as food frequency questionnaires (FFQ). Objective: The objective of this study was to validate e-EPIDEMIOLOGY as a tool for the determination of habitual intake of certain foods/drinks, using a traditional FFQ as a reference method. Methods: University students between the ages of 18 and 24 years of age recorded the consumption of certain foods/drinks using an application for mobile phones (e-EPIDEMIOLOGY) during 28 consecutive days and then filled out a conventional FFQ on paper at the end of the study period. The agreement between the category of habitual consumption for each of the foods/drinks included in the study was evaluated using cross-classification analysis and average weighted kappa statistic. Results: 75 participants completed the study (23% male and 77% female). Cross-classification analysis showed that more than 69% of the participants were correctly classified into the same category and less than 3% were misclassified into an opposite category. The average weighted kappa statistic was moderate (k = 0.43). Conclusions: e-EPIDEMIOLOGY could be considered a reasonably and moderately valid tool in epidemiological studies to measure single foods/drinks habitual intake in young adults, as a valuable alternative to conventional paper FFQ. Objective: Introduction Characterization of the intake of foods/drinks has been used in the last few decades in numerous epidemiological studies, though with different objectives: 1) In descriptive studies, as a basis for the planification of certain Public Health policies related to diet; 2) In analytical observational studies, in order to define relationships between the consumption of certain foods/drinks and disease; and 3) In analytical experimental studies that evaluate Public Health policies whose intent is to modify certain concrete dietary habits [1-4]. Traditional methods that evaluate dietary intake, such as dietary registries and 24 hour recall questionnaires (short-term methods) and FFQ (long-term instruments) present important limitations. The short-term tools allow collection of data that includes quantities of all of the foods/drinks consumed by a person during a certain number of days. Dietary registries that require weighing of foods are time-consuming and suppose a great deal of work for study participants, which can lead to deviations from normal food intake (especially underestimation of quantities), as well as low rates of participation and compliance. Dietary registries and 24 hour recall questionnaires also require trained personnel and are short-term memory dependent [5-8]. It is also important to remember that in order to determine habitual dietetic intake (the long-term mean consumption of foods) using these short-term tools, it would be necessary to repeat multiple times, which would only worsen the problem. Long-term recall methods allow information to be collected about the consumption of a series of foods/drinks during prolonged periods of time (weeks or months), classifying a person according to the consumption category applied to each of the foods/drinks considered. FFQ depend mostly on the memory of the subject interviewed, do not allow for intrapersonal variation in the recording of daily food consumption during the time period of the study and do not allow precise estimation of food portion size. Both long-term and short-term tools if use conventional techniques (paper and pen) to collect information, with posterior manual introduction for statistical analysis, which increments research costs and time consumption considerably [3,7]. For these reasons, improvement upon traditional methods for the determination of dietary intake still remains one of the most important challenges in nutritional epidemiology [2-4,9]. Improvement of self-reporting that contributes to greater precision of habitual dietary intake would represent a considerable boon for researchers, as well as for society as a whole, keeping in mind the important repercussions that the results and conclusions of these studies can have on the general population. Traditional techniques that evaluate dietary intake should be substituted by new solutions, or nutritional research and treatments for nutritional problems will remain restricted and deficient . Recently certain dietary registries and 24-hour recall mobile phone applications have been developed that could reduce the limitations of these methods, with promising results [1-4,11]. However, and though FFQ are the most practical, accessible and commonly utilized tools in research that determines habitual dietary intake [6,9,12], up until now, no long-term instrument has been developed that takes advantage of mobile technology and serves as an alternative to traditional FFQ. Use of mobile phones is extense in Spain, with 95.0 % of Spaniards having used a mobile phone in the last 3 months . This facilitates the introduction of new methods of evaluation of dietary intake that include mobile technology. In any case, these new technologies need to be developed according to different local conditions and evaluated with objective measures . Our research team has developed a new application for mobile phones called e-EPIDEMIOLOGY, designed to collect individual consumption data of a series of foods/drinks. The objective of this study was to compare data recorded with e-EPIDEMIOLOGY with that registered through a conventional paper FFQ for the same foods/drinks, in order to evaluate its potential as research tool for the determination of habitual dietary intake. Methods: Methods Study Sample This study was performed among medical and pharmaceutical students of the University of Seville (Andalusia, Spain, Southern Europe). Different events were programmed at both faculties in which the research team personally presented the project to the students. At the end of each presentation, interested students and those that fit inclusion criteria signed up for a personal interview. Of the 127 students that were interested, 88 were eligible and were signed up for the interview, in which the study protocol was explained. Finally 76 students decided to participate in the validation study. Of these, 75 completed both the application e-EPIDEMIOLOGY and the conventional paper FFQ. The period of participant recruitment spanned from October 2014 to January 2016. The inclusion criteria were: University of Seville student from the medical or pharmaceutical schools, between the ages of 18-24 years old and the owner of a mobile phone with access to the Internet and an Android operating system. As an incentive, all participants were entered into a raffle of a tablet at the completion of the study. The study was performed according to directives established in the Helsinki Declaration and the Biomedical Research Law , and all procedures on human beings were approved by the Ethical Committee for Experimentation of the University of Seville. Written informed consent was obtained from all participants. e-EPIDEMIOLOGY, a mobile phone application Participants downloaded the application e-EPIDEMIOLOGY to their personal mobile phones. This application was developed for mobile phones with an Android operating system and permitted the recording of daily consumption of the foods/drinks selected for the study. At the end of each day a notice would appear on the participants mobile phone, informing them that it was time to use the application. At this moment, the participant could access the application and register the number of standard portions that had been consumed during that day of each of the foods/drinks included in the study. The list of foods appeared every day in the same order to facilitate completion of the application. This list consisted of 12 items which referred to 10 different foods/drinks: fruit, vegetables, red meat (lamb, beef, and pork), chicken/turkey, fish, legumes, sweets, prepared foods, soft drinks and alcoholic beverages. These were selected for the study because they provide consumption patterns that range from daily to sporadic for the population . These were also considered to be markers for healthy (fruits, vegetables, fish and legumes) and unhealthy (sweets, prepared foods, and soft drinks) dietary habits . When accessing the first food/drink on the list, the number of standard portions of this food/drink consumed throughout the day were introduced. The button 'Next' was then pressed to go on to the next item in order to record all foods/drinks consumed that day (Figure 1). One could correct errors by pressing 'Go Back' and changing the number introduced. After filling out e-EPIDEMIOLOGY, the data was automatically saved and sent to the research administrator's website via Wi-Fi or 3G/4G, after which time the user could not access or change answers to the questionnaire. This process was repeated every day for each user during 28 days. The time necessary to complete e-EPIDEMIOLOGY was about 1 minute per day. Each participant selected the time of day at which the reminder would be set, during the interval between 8 pm and midnight (after having consumed all of the foods and beverages for that day). The application was blocked at midnight until the next day at 8 pm. Saved data could be consulted pressing the 'Historial' button, though this information could not be changed or eliminated. Figure 1. Screen captures of the application e-EPIDEMIOLOGY. The application used to register daily consumption of selected foods/drinks was based on a questionnaire elaborated using the FFQ from the European Health Survey  (Appendix 1). Standardized portions were added after testing a previous prototype of e-EPIDEMIOLOGY (results not published) and were obtained from a FFQ validated for the Spanish population . The application also allows for registry of other lifestyle habits (hours of sleep, oral hygiene, physical activity and tobacco consumption). The application recorded this information using a different questionnaire with 11 items, also based on validated instruments from the European Health Survey . Appendix 1. Questionnaire used in e-EPIDEMIOLOGY, with weights / measurements of standardized rations of selected foods/drinks. 1. How many pieces of fruit have you eaten today? (1 piece = aprox. 100 g) (Include fresh-squeezed juice (1 ration = aprox. 200 ml)) 2. How many portions of vegetables have you eaten today? (1 portion = aprox. 150 g) 3. How many portions of legumes (lentils, garbanzos, beans, etc.) have you eaten today? (1 portion = aprox. 60 g) 4. How many portions of chicken/turkey have you eaten today? (1 portion = aprox. 150 g) 5. How many portions of fish have you eaten today? (1 portion = aprox. 150 g) 6. How many portions of red meat (beef, pork, lamb) have you eaten today? (1 portion = aprox. 150 g) 7. How many servings of soft drinks have you had today? (1 serving = aprox. 250 ml) 8. How many portions of comercially produced sweets (not home-made) (cookies/pastries) have you eaten today? (1 piece = aprox. 100 g) 9. How many portions of prepared/frozen foods have you eaten today (croquettes, pizza, etc.)? (1 portion = aprox. 80 g) 10. Have you consumed alcoholic beverages today? 11. What kind of alcoholic beverage have you consumed? 12. How many servings of beer/wine/spirits or mixed drinks have you consumed today? (1 serving of beer = aprox. 200 ml / 1 glass of wine = aprox. 100 ml / 1 serving of spirits or mixed drinks = aprox. 50 ml (of alcohol)) Anthropomorphic measurements Researchers used the personal interview to both explain the study protocol and collect anthropomorphic data using standard procedure. Height was measured in centimeters, with a precision of 0.5 cm, and weight in kilograms, with a precision of 0.1 kg (wearing lightweight clothing, with shoes off and pockets empty). With this data, BMI (Body Mass Index) (Kg/m2), was calculated using categories defined by the WHO . Procedure All participants completed a questionnaire during the personal interview in which demographic data was collected, such as date of birth, gender, birthplace, and current place of residence. Participants were instructed in the use of e-EPIDEMIOLOGY with a personal demonstration of use of the application, as well as estimation of standardized portion sizes, and were reminded to maintain their habitual diet. The recording of foods/drinks intake was to be completed during 28 consecutive days using the application. Participants were recruited to the study during the entire period of research, so that all seasons, days of the month and week were included in the sample. As a reference, a conventional paper FFQ was filled out at the end of each period of the study, through personal interviews and at the convenience of the participants. The FFQ utilized was based on a validated questionnaire used in the European Health Survey (Appendix 2) . Standardized portion sizes were obtained from an FFQ validated for the Spanish population . Both the questionnaires used in the application and the paper FFQ had the same items (Appendixes 1 y 2), the only difference being that in e-EPIDEMIOLOGY the questionnaire refers to daily consumption while the FFQ refers to consumption during the previous 28 days. Appendix 2. Questionnaire utilized for conventional paper FFQ, with weights / measurements of standardized portions of selected foods/drinks. 1. How many pieces of fruit did you habitually consume in the last 28 days? (1 piece = aprox. 100 g) (Include fresh-squeezed juice (1 portion = aprox. 200 ml)) Categories a 2. How many portions of vegetables did you habitually consume in the last 28 days? (1 portion = aprox. 150 g) Categories a 3. How many portions of legumes (lentils, garbanzos, beans, etc.) did you habitually consume in the last 28 days? (1 portion = aprox. 60 g) Categories a 4. How many portions of chicken/turkey did you habitually consume in the last 28 days? (1 portion = aprox. 150 g) Categories a 5. How many portions of fish did you habitually consume in the last 28 days? (1 portion = aprox. 150 g) Categories a 6. How many portions of red meat (beef, pork, lamb) did you habitually consume in the last 28 days? (1 portion = aprox. 150 g) Categories a 7. How many servings of soft drinks did you habitually consume in the last 28 days? (1 serving = aprox. 250 ml) Categories a 8. How many portions of comercially produced sweets (not home-made) (cookies/pastries) did you habitually consume in the last 28 days? (1 piece = aprox. 100 g) Categories a 9. How many portions of prepared/frozen foods have you habitually eaten (croquettes, pizza, etc.) in the last 28 days? (1 portion = aprox. 80 g) Categories a 10. Have you consumed alcoholic beverages in the last 28 days? Yes No 11. What kind of alcoholic beverages have you consumed in the last 28 days? Beer Wine Spirits/mixed drinks Others 12. How many servings of beer/wine/spirits or mixed drinks did you consume in the last 28 days? (1 serving of beer = aprox. 200 ml / 1 glass of wine = aprox. 100 ml/ 1 serving of spirits or mixed drinks = aprox. 50 ml (of alcohol)) Categories a a The different categories were: : Less than once a week / Once or twice a week / 3-4 times a week / 5-6 times a week / Once or twice a day / 3 times or more a day All of the personal data collected in this study remained anonymous and confidential and were treated according to the current Spanish legislation . To that end, each participant was assigned a personal alphanumeric code, so that no-one, including the researchers, could link personal information to the results obtained. The code was introduced the first time the participant accessed the application, and when completing the demographic questionnaire and paper FFQ, for organizational purposes. Codification y revision of data For each participant, the data collected from the FFQ for each of the 10 foods/drinks mentioned previously were categorized. The frequency of consumption of foods/drinks ítems was categorized into six subgroups, ranging from “Less than once a week” to “3 times or more a day” (Appendix 2). For the same foods/drinks, the data from the 28 days using e-EPIDEMIOLOGY were recorded as daily consumption. This data was transformed in order to include it in one of the same categories of habitual consumption included in the FFQ. This was made possible because both the FFQ and e-EPIDEMIOLOGY used the same standardized portion sizes. For example, suppose that a participant consumes an average of 0.25 standard rations of fish daily during 28 days using e-EPIDEMIOLOGY. This average consumption represents 1.75 standard portions per week (0.25 x 7 = 1.75), which would be classified in the category “Once or twice a week.” The data collected from the conventional paper FFQ were manually introduced in the data base by the research team. These were reviewed in order to avoid data entry errors. Data collected from e-EPIDEMIOLOGY were saved without modifications in a separate data base. Posteriorly, one set of data was removed due to obvious inconsistency: one participant registered the consumption of 200 standardized portions of legumes in one day. Statistical analysis Due to the lack of agreement on the best way of presenting results from validation studies, it is necessary to use more than one statistical method in order to give credence to the results. In this study cross-classification analysis and the weighted kappa statistic were used. e-EPIDEMIOLOGY and conventional FFQ are designed to rank individuals rather than to assess their absolute level of intake, thus a correlation coefficient would not apply. To assess agreement, subjects were classified into categories of intake by e-EPIDEMIOLOGY and the reference method, and the percentage of subjects correctly classified into the same category and grossly misclassified into the opposite category were calculated. With cross-classification, the percentages misclassified clearly illustrate the likely impact of measurement error; however, the percentage of agreement will include agreement that can be accounted for by chance. Weighted kappa statistic is a summary measure of cross-classification that allows for the agreement expected by chance and has the added advantage over the kappa statistic in that it allows for the degree of misclassification. However, both the cross-classification analysis and the weighted kappa statistic are still dependent on the number of categories used. In order to limit this dependence, the six original categories were reorganized into three (Category 1: “Less than once a week” and “Once or twice a week”; Category 2: “3-4 times a week” and “5-6 times a week”; Category: “Once or twice a day” and “3 times or more a day”), in order to apply Masson and colleagues criteria  to evaluate agreement and misclassification. The inter-rater agreement of two assessment methods was analysed by weighted kappa statistic , assigning partial credit to scores using the Stata prerecorded weights. If there was complete agreement, a weight of 1.00 was assigned. Slight disagreements (off by one) were given a weight of 0.50 and 0.00 if there was a complete disagreement. Values of kappa over 0.80 indicate very good agreement, between 0.61 and 0.80 good agreement, 0.41-060 moderate agreement, 0.21-0.40 fair agreement and < 0.20 poor agreement . All statistical analysis was performed using STATA version MP 13.1 (Stata Corp LP, Texas, USA) and a P value <.05 was considered statistically significant . Results: Results 76 individuals participated in the study, but one participant did not finish neither the application nor the FFQ. This individual’s data was not used for posterior analysis. The study potentially could have generated 2128 separate data entries using e-EPIDEMIOLOGY (76 participants x 28 días = 2128). 2054 separate data entries were obtained (96.5%), meaning the application was completed on 2054 days and not completed on 74 days (28 of which are due to the participant mentioned previously that did not complete any days of the application). Of the rest of the participants 56 individuals (74.7%) completed the application every day and the remaining 19 filled out the application at least 24 of the 28 days. Among the participants, the mean age was 21.0 years. 22.7% were males and 77.3% were females. 14.7% were smokers. About categories of BMI (Kg/m2): 4.0%, underweight; 77.3%, normal range; 14.7%, overweight; 4.0%, obesity (Table 1). The percentage of individuals correctly classified into the same category ranged from 53% (vegetables) to 80% (legumes), while the percentage of individuals misclassified into an opposite category ranged from 0% (red meat, soft drinks and prepared foods) to 7% (sweets) (Table 2). Weighted kappa statistic values showed good agreement for fruit (k = 0.61), moderate agreement for chicken/turkey, soft drinks and alcoholic beverages (k = 0.53-0.57) and fair agreement for vegetables, legumes, fish, red meat, sweets and prepared foods (k = 0.21-0.40) (Table 3). Table 1. Characteristics of participants in the study. Age, years, mean (SD) 21.0 (1.7) Gender, N (%) Male 17 (22.7) Female 58 (77.3) Smoking status, N (%) No 64 (85.3) Yes 11 (14.7) BMIa, kg/m2, N (%) Underweight 3 (4.0) Normal range 58 (77.3) Overweight 11 (14.7) Obesity 3 (4.0) a BMI: body mass index Table 2. Cross-classification analysis derived from e-EPIDEMIOLOGY and conventional paper FFQ. Comparison Agreement (%) Same category Adjacent category Extreme category Fruit 72.0 22.7 5.3 Vegetables 53.3 42.7 4.0 Legumes 80.0 18.7 1.3 Chicken/turkey 72.0 22.7 5.3 Fish 72.0 26.7 1.3 Red meat 70.7 29.3 0.0 Soft drinks 70.7 29.3 0.0 Sweets 54.7 38.7 6.7 Prepared foods 74.7 25.3 0.0 Alcoholic beverages 77.3 21.3 1.3 Average 69.7 27.7 2.5 Table 3. Percentage agreement, percentage expected agreement and weighted kappa statistic derived from e-EPIDEMIOLOGY and conventional paper FFQ. Comparison Agreement (%) Expected agreement (%) Weighted kappa P Fruit 83.3 56.9 0.61 < 0.01 Vegetables 74.7 57.5 0.40 < 0.01 Legumes 89.3 83.5 0.34 0.01 Chicken/turkey 83.3 64.6 0.53 < 0.01 Fish 85.3 76.8 0.37 < 0.01 Red meat 85.3 75.8 0.39 < 0.01 Soft drinks 85.3 66.1 0.57 < 0.01 Sweets 74.0 59.4 0.36 < 0.01 Prepared foods 87.3 83.8 0.21 0.02 Alcoholic beverages 88.0 73.8 0.54 < 0.01 Average - - 0.43 - Conclusions: Discussion Principal Findings This is the first study that evaluates an alternative to traditional FFQ using mobile technologies. Recently, certain dietary registries and recall questionnaires that use mobile technologies have been developed, with promising results [1-4,11]. However, until now, none have been developed that evaluate long term intake, as well as benefit from mobile technologies and serve as alternatives to the conventional paper FFQ. An accurate method is one that measures what the method intends to measure, i.e. the “truth”. In the context of dietary studies “the truth” represents actual intake over the period of the study . However, there is not, and probably never will be, a method that can estimate dietary intake without error . The semi-quantitative FFQ is the primary dietary assessment method used in epidemiological studies . Results from such studies can be interpreted with greater confidence if the questionnaire has a quantified validity. To assess the true validity of a FFQ would require measuring with high accuracy the usual self-selected diet of free-living individual over several months, which is not feasible. Therefore, research assesses relative validity by comparing the FFQ with an alternative dietary assessment method with its own limitations [6,20]. In epidemiological studies, the odds ratio or relative risk of disease in relation to nutrient intake is the most common measure association presented. Consequently, FFQ must be able to rank individuals along the distribution of intake, so that individuals with low intakes can be separated from those with high intakes. Therefore, obtaining absolute nutrient intakes is not necessary. As long as FFQ can rank individuals, relative risk estimates will be accurate . It is also unnecessary to record exact consumption of nutrients in those descriptive studies which aim to help plan concrete Public Health measures related to diet, or in analytical experimental studies which aim to help create policies that will modify certain dietary habits. In the first case, one would only need to record the amounts consumed of certain kinds of foods, while in the second, it would be necessary to measure the change in consumption over time. The fact that there is error in dietary measurements does not mean that dietary data should not be collected but simply that it is important to determine the nature of the errors associated with dietary data so that these can be taken into account in evaluating the data . In this study the relative validity of e-EPIDEMIOLOGY was determined using a previously validated conventional paper FFQ . Often in previous studies, long-term methods such as an FFQ were compared to short-term methods, such as 24 hour recalls, each of these presenting their own limitations. It would seem more interesting to compare the results of two long-term instruments that allow categorization of individuals based on their consumption of certain foods/drinks, without recording total consumption. This would allow analysis of the capability of e-EPIDEMIOLOGY to minimize some of the limitations that, in theory, are shared with conventional paper FFQ. The objective of this study was to compare data collected with e-EPIDEMIOLOGY with registries obtained with conventional paper FFQ and evaluate its potential as a tool for the determination of habitual dietary intake in research. Cross-classification analysis showed that more than 69% of the participants were correctly classified into the same category and less than 3% were misclassified into an opposite category, which depicted good agreement and lower misclassification between two methods, thus demonstrating that e-EPIDEMIOLOGY is able to ranking subjects on a range of nutrient intakes. Presenting data categorized in categories provides compact information concerning the capacity of both methods to allocate individuals according to dietary intake distribution . The average weighted kappa statistic was moderate (k = 0.43), with values over 0.35 in 8 of the 10 foods/drinks selected for the study. Thus, due to the moderate agreement with conventional paper FFQ estimated by the mean weighted kappa statistics, as well as its good ability to classify individuals into categories, e-EPIDEMIOLOGY could be considered a reasonable and moderately valid instrument for correctly ranking subjects into classes of food groups, according to Masson’s criteria . FFQ have been demonstrated to be valid for ranking subjects on a range of food intakes even though there is an on-going need for the refinement of these tools . Though e-EPIDEMIOLOGY can be considered a valid tool for the classification of individuals into categories of habitual consumption of foods/drinks, the results of this study show that there is obvious disagreement between both instruments (cross-classification analysis showed that 28% of the participants were incorrectly classified into an adjacent category and 2.5% were misclassified into an opposite category). Misclassification of participants could have important negative on the results and conclusions of studies that use such tools. The analysis of the characteristics of each of these methods shows that e-EPIDEMIOLOGY is a more precise and more refined method than the traditional paper FFQ for the correct classification of individuals into categories of habitual consumption. This held true when there was not agreement between both methods. In terms of precision, both methods have in common that, for each of the foods/drinks considered, both use the same question to measure the frequency of consumption. For example, both ask: 'How many portions of fish have you eaten? (1 portion = aprox. 150 g)' The difference between both methods lies in that e-EPIDEMIOLOGY this question is answered at the end of each day during the study period, while the FFQ is completed at the end of 28 days. Consequently, both methods present the same difficulties in the precise estimation of portion size, given that standardized ration size are used in both. However, e-EPIDEMIOLOGY permits daily collection of information, FFQ only allows collection of information at the end of the study period. This minimizes the dependence on the memory of the participant in e-EPIDEMIOLOGY in comparison to the FFQ, keeping in mind that the recollection of past consumption of foods can be influenced by more recent food consumption . e-EPIDEMIOLOGY allows for daily intrapersonal variability in the collection of consumption of foods/drinks. Among university students, who made up our study sample, dietary intake is variable from day to day, with sporadic changes in food intake (skipping meals, snacking, school events that interfere with meal time), as well as frequent dining out. These aspects interfere with the precise determination of habitual dietary intake , especially in the case of FFQ, where data is collected only once at the end of an extended time period. Repeated applications of traditional short term instruments, such as dietary registries and 24 hour recalls, can modify habitual intake due to the excessive work load for participants. Any tool that aports a simple method that facilitates data collection about dietary intake without changing behavior is an important advance . Despite repeated use, the modification of habitual intake seems unlikely through the use of e-EPIDEMIOLOGY, due to the reduced workload that using this application presents (1 minute/day). In relation to refinement, one must consider other aspects related to its use, such as research costs and the ease of digitalization of data. If information is recorded in a traditional manner (pen and paper), such as with the FFQ used in this study, the costs increase due to need for interviewers and for the digitalization of data for posterior statistical analysis, which also increases time consumption [3,7]. The use of mobile technologies as tools for self-reporting (for example, e-EPIDEMIOLOGY) can eliminate the need for interviewers, and permits the instant digitalization of data [2,15,26], minimizing costs and facilitating research. If this study had compared the application e-EPIDEMIOLOGY with a FFQ applied through a mobile phone application, those aspects related to cost and digitalization of data would have been equalized though this would not have affected the precision of the FFQ, due to the fact that this does not depend on the format used, but on other considerations aforementioned. In this study a new instrument, using mobile technology (e-EPIDEMIOLOGY), was compared with the more frequently used paper FFQ, for this reason an electronic FFQ (website or mobile phone application) was not used. FFQ are tools that allow collection of data related to consumption of foods/drinks, including those consumed sporadically, in one use. This characteristic would seem to make it advantageous compared to e-EPIDEMIOLOGY in terms of the time necessary to obtain information from all of the participants in the study. The FFQ only requires a few minutes to complete, while e-EPIDEMIOLOGY requires completion every day for 28 days. From a practical point of view, this aspect does not suppose a great difference for the researcher. For example, in the current study the participants were recruited between October of 2014 and January of 2016. If an FFQ were used, the information from the 76 participants would be available in January 2016, when the last participant had completed the questionnaire. If only e-EPIDEMIOLOGY had been applied, the data from the 76 participants would be available in February 2016, the last day of completion of the application. In a study were recruitment occurred during 16 months, this difference of one month seems irrelevant. However, the application also has the many advantages mentioned previously and even allows analysis of the participant's behavior during different days of the week, during vacations, etc. Another advantage is that the investigators receive the data directly to their computer. Limitations One of the possible limitations of this study is the possible rate of non-response. On the one hand, those who did not fill out any days of e-EPIDEMIOLOGY and, on the other, those who did not fill out the application during one or a few days of the 28 days duration of the study. Of the 76 participants in the study, 56 completed the application every day (74.7%), 19 completed the application at least 24 of the 28 days of the study and just one individual did not complete any of the 28 days. Some of the characteristics of these types of mobile technologies, such as asinchrony [27-30], the ease with which privacy can be maintained , as well as the light workload for the participants (1 minute per day), helped to increase participation and could have contributed to minimize the rate of non-response. Young people have expressed their preference for those methods of evaluation of dietary intake that utilize new technologies, as they can easily be incorporated into their lifestyles and are more amenable than traditional pen and paper methods [1,2]. In any case, the possible limitation presented by the rate of non-response was minimized, as no statistically significant differences were found in any of the variables studied (age, gender, tobacco consumption, BMI), after analysing the basic characteristics of responders and non-responders. Another possible limitation lies in the fact that access to these technologies is not universal, excluding especially vulnerable groups, such as students from poorer social strata. In the environment in which this study was performed, the percentage of students with mobile phones with internet access is very high, which minimizes this possible limitation. Future Studies To further study the potential of e-EPIDEMIOOGY as a tool for the evaluation of habitual dietary intake, the research team intends to perform future investigations in different socio-demographic groups in order to increase the representativity of the results and conclusions obtained, of the general population. As no similar studies were found in the literature, we could not compare with previous experiences in order to help select a follow-up time, so the duration of 28 days was chosen because it fit the characteristics and objectives of this study. We would like to evaluate e-EPIDEMIOLOGY, modifying the follow-up time, reducing from daily data input to input 2-3 times a week, as well as varying the foods/drinks selected. It is worth noting that another advantage of these technologies is that the applications are easily modifiable (varying the questions, standardized portions, time of completion of the application, reminder time, etc.) permitting adaptation to different sociocultural idiosyncrasies present worldwide. Another line of study would be to analyse the impact of factors that can affect the validity of data collected with e-EPIDEMIOLOGY, such as age, gender, IMC and health related behavior (level of physical activity, tobacco consumption, etc.). Due to the small sample size, separate statistical analysis was not performed for males and females, nor for distinct categories of BMI, physical activity, etc. In all of these future projects, the development of the process will include estimation of portion size using digital images in order to decrease error. Conclusions e-EPIDEMIOLOGY could be considered a reasonably and moderately valid tool in epidemiological studies to measure habitual intake of certain food/drinks in young adults, as a valuable alternative to conventional paper FFQ. Along with the growing popularity of mobile phones among young adults, this instrument is likely to be accepted in this population and could reduce some of the inherent limitations present in paper FFQ, such as dependence on the memory of participants and the impossibility to reflect intrapersonal variability in daily consumption of foods/drinks. Future studies should aim to explore the validity of e-EPIDEMIOLOGY in different samples, modifying foods and drinks analysed, duration of the study, etc. This could confirm its value as a tool to determine habitual dietary intake in both descriptive and analytical epidemiological studies (both observational and experimental). Acknowledgements We would like to thank the participants in the validation study. This research was partly supported by funding from the Research Plan of the University of Seville. Authors' Contributions LMB performed the conception and design of the study, developed the application, analysed and interpreted data, and wrote the article; BS and MDG were involved in data collection and interpretation of the data and contributed in drafting the article; and all authors were involved in editing the final draft of the article and revising it critically and approving the manuscript. Conflicts of Interest None declared.
A Review of Safe Sex Messages within Smartphone Applications
Date Submitted: Mar 13, 2016
Open Peer Review Period: Mar 14, 2016 - May 10, 2016
Background: Smartphone applications provide a new platform for entertainment, information distribution and health promotion activities, as well as for dating and casual sexual encounters. Previous res...
Background: Smartphone applications provide a new platform for entertainment, information distribution and health promotion activities, as well as for dating and casual sexual encounters. Previous research has shown high acceptability of sexual health interventions via smartphone apps, however, sexual health promotion apps were infrequently downloaded and underutilized. Therefore, using established apps to integrate sexual health promotion might be a more effective method. Objective: To critically review popular sex-related apps and dating apps, in order to ascertain if there are currently any sexual health messages present in the apps. Methods: Part 1: We used the search term “sexual” to search for free apps in Apple iTunes store and Android Google Play store, and analyzed sexual health content on the 137 apps included. Part 2: The search term “dating” was used to search for free geosocial networking apps in the Apple iTunes and Android Google Play stores. The apps were downloaded to test out functionality, and to search for sexual health messages therein. Results: Part 1: Of the included 137 apps, 15 (11%) contained sexual health messages, and 14 (10%) contained messages for sexual assault/violence. The majority of the apps did not contain any sexual health messages. Part 2: Sixty dating apps were reviewed: 44 (73.3%) apps targeting heterosexual users, 9 (15%) apps targeting men who have sex with men (MSM), 3 (5%) apps targeting lesbian women, and 4 (6.7%) apps for group dating. Only 9 dating apps contained sexual health messages, of which 7 targeted MSM. Conclusions: The majority of sex-related apps and dating apps contained no sexual health messages that could educate and remind users of their sexual risks. Sexual health practitioners and public health departments will need to work with app developers to promote sexual health within existing popular apps. For those apps that already contain sexual health messages, further study to investigate the effectiveness of the messages is needed.