Development of the TabacoQuest app for computerization of data collection on smoking in psychiatric nursing

ABSTRACT Objective: to develop a mobile app for research on the use of tobacco among psychiatric patients and the general population. Method: applied research with the technological development of an app for data collection on an Android tablet. For its development, we considered three criteria: data security, benefits for participants and optimization of the time of researchers. We performed tests with twenty fictitious participants and a final test with six pilots. Results: the app collects data, stores them in the database of the tablet and export then to an Excel spreadsheet. Resources: calculator, stopwatch, offline operation, branching logic, field validation and automatic tabulation. Conclusion: the app prevents human error, increases the quality of the data by validating them during the interview, allows the performing of automatic tabulation and makes the interviews less tiring. Its success may encourage the use of this and other computational resources by nurses as a research tool.

The rapid spread of mobile devices in the market has favored the integration of this technology in the area of health, both in research studies and in care assistance and management. In a Swedish study, with 398 nurses and nursing students, it was found that most believed that mobile devices can benefit nursing efforts, without prejudice to the quality of the care and increasing the confidence of patients in the professionals (6) . It is believed that this increased confidence can be extended to researchers.
In addition to the benefits to the respondent, the computerized data collection decreases the possibility of human error during registration of the answers, contributing to a better quality and security of the data and increasing the scientific credibility of the research results (3,(7)(8)(9) .
The MCAPI has been introduced in epidemiological Although we found no studies on the development of this type of app for data collection, on the use of tobacco among psychiatric patients, we believe that it is relevant because it is more dynamic, attractive and interactive than paper-based research studies.
This study aimed to develop a mobile app for research on the use of tobacco among psychiatric patients and the general population.

Methods
This is an applied research with the technological development of a app for computerized data collection  We decided to develop an app for mobile devices, named TabacoQuest, using the Android operating system as it is a platform distributed widely and Oliveira RM, Duarte AF, Alves D, Furegato ARF.

Results
The results are presented in three topics: A) presentation of the app , B) main resources, purposes and advantages and C) limitations of the app.

Presentation of the app
In order to make the app attractive to the research participants, different interface options were contemplated. We eliminated the possibility of background with light color because of higher battery consumption. We opted for the interface that best highlighted the text (dark blue background), in addition to being attractive ( Figure 2). We used Arial Rounded MT Bold, in yellow for the questions and words requiring emphasis and in white with shadow effect for the answers. In Figure 2 we show some TabacoQuest screenshots. The initial screen of the app has functionality that automatically records the number of the questionnaire and the date of the interview (Figure 2a). On this screen, the nurse/researcher must enter the location of the study and select the "next" button to be directed to the following screen.
We conceived and built a user-friendly interface, touches in a single screen ( Figure 2). The answers were all filled by the same nurse; however, in order to promote interaction with the interviewees, in some moments, they would be encouraged, with the supervision of the interviewer, to fill their own answer by touching the screen of the tablet (Figure 2c and 2e). The graphics were attractive and enabled interaction, thus providing better attention and motivation to the participants.
On the last screen of the app, a message appears indicating that the interview is done (Figure 2f). The nurse selects in the menu the option "save form" so that the answers can be stored in the native database of the tablet. To start a new interview, "new form" needs to be selected. To display the data, we need to select the option "export database" and all data stored on the tablet will be exported to an Excel spreadsheet. The connection is performed via an USB cable to the computer and then we can recover the file in Excel format, containing the data.

Main resources, purposes and advantages
In Figure 3, we show some TabacoQuest screenshots that exemplify the resources in it. In order to avoid blank answers, the field validation feature has been inserted ( Figure 3a and 3c). The app does not allow the interview to go on if any variable is not answered, thus ensuring the consistency and integrity of the data that will be stored. A red alert is displayed, indicating the need to select an alternative.
The questions are presented in the app according to the branching logic. In this way, the questions that are not applied to a particular participant are filled automatically or, in some cases, omitted. Figure 3b illustrates that, when the participant notes that he or she has no children, the next question (number of children) is registered automatically as zero. The same occurs with the variable "self-reported religion". If the participant has pointed out no religion, the variable "religious practice" is automatically registered as "not applicable".
Initially, we planned to apply the branching logic to all questionnaires. By examining the criteria for inclusion of resources, we admitted that although the branching logic could benefit participants (less tiring interviews) and optimize the time of the researcher, data security could not be guaranteed if this resource was applied in PETab questionnaires.
The PETab is a complex questionnaire which presents conditions for all the answers (for example, if the answer is "daily" for question D1, we skip to question D4; if the answer was "less than daily" for question D1, we move on to question D2; if the answer is "never smoked" for question D1, we skip to question D3). We considered that it would not be safe the automatic guidance of the questions, which would then be conducted manually by the nurse/researcher. In order to guide the respondent through the interviews, statements have been inserted in front of each PETab answer, as well as shortcuts at the top of the app screen ( Figure 3c). With this, if the nurse marks "daily" to question D1, there is a statement that the next question that must be answered is D4. The nurse/researcher clicks the shortcut "D4" at the top and is directed to the screen related to this question. In this case, the unanswered questions (D2 and D3) are automatically registered as "not applicable" in the database.
In Figure 3d and D19 must be recorded in the database as "days".
However, the nurse/researcher has the option of inserting the answer as the participant gives them into the app.
If the answer is in months, the number of months is inserted in the text box with the guidance "months". In the database, the answer will appear as "days".
(the Figure 3 continue in the next page...) Oliveira RM, Duarte AF, Alves D, Furegato ARF. In addition to the benefits already mentioned, we also saw little difference in the cost that the researchers would have with the printing of the questionnaires and the cost of the tablet.

Limitations of the app
The main limitation of the app refers to its one-way operation. When initiated a new interview, the previous answers can no longer be retrieved (displayed) on the screen of the app. If there is the need to change any information, after the interview is saved, the researcher must perform it manually in the Excel spreadsheet.
When an interview is interrupted, the researcher does

Discussion
When we thought about the possibility of a computerized data collection for research on smoking, we performed a search on the tools available. Although several options have been found, most required Internet connection at the time of the interview, were paid or did not have the resources necessary to address the complexity of some questionnaires used in the research. Therefore, we decided to develop an app.
For its development, we considered three criteria: 1) data security, 2) benefits for the participants Studies carried out in the United States (n=49), India (n=95) and Fiji (n=120) show that participants, including older ones, prefer to answer computerized questionnaires than answer printed questionnaires. For them, paper research requires more time, besides not being dynamic (3,8,14) .
A Swedish study has shown one of the main concerns regarding the use of mobile devices in the health field: the loss of the professional-patient interaction (6) . Similar to what occurred with the use of TabacoQuest, by reporting their experience with the use of Palm OS, for computerized data collection, a researcher said that the use of the mobile device favored the interaction with the patient and allowed greater eye contact than interviews conducted in paper (4) . The favoring of the visual contact was also reported in a Chinese study (15) .
Besides facilitating the visual contact between interviewer/interviewee and increasing the interest of persons in participating in the research, the computerized data collection has other advantages, depending on the resources that are included in the app: 1) data are validated at the time of the interview (which eliminates the possibility of blank answers or wrong selection of more than one answer), 2) it is not necessary to manually tabulate the data, thus decreasing potential human errors, 3) long questionnaires become less tiring (only the questions that apply to each participant are presented) and 4) data analysis can be started immediately after completing the number of participants, as it is not necessary to tabulate them (3,8) .  Although the limitations of TabacoQuest have not harmed the research, we hope to develop alternatives to fix them. Future studies can compare the application of questionnaires in paper and using the tablet, in order to assess the acceptance of the app by the participants.

Conclusions
The TabacoQuest