An Overview of the Research Reports on Addiction in Iranian Journals from 2010 to 2015.

BACKGROUND: The articles and research reports presented in the scientific journals are selected without a proper critical appraisal in many cases. This is a major problem, as it leads to the opinions of the experts being neglected in many of these scientific journals. A descriptive study was done in fall and winter of 2016.
METHODS: Journals were identified searching the electronic databases, including PubMed, Medline, Scopus, Google Scholar, and national Persian databases of SID and Magiran. All the articles on addiction published in the time span of 2010-2015 in 8 psychiatry, psychology, and addiction journals were included in this study. The journals included Journal of Behavioral Sciences, Iranian Journal of Psychiatry and Clinical Psychology, Journal of Fundamentals of Mental Health, and Journal of Research on Addiction in Persian, and Health and Addiction Journal, International Journal of Psychiatry and Behavioral Sciences, Journal of Research in Behavioral Sciences, and International Journal of Psychological Studies in English. These journals were evaluated according to their research design and methodology, scientific writing style, and the validity of the references. For all of the 264 articles, the research design and references were thoroughly evaluated.
FINDINGS: The most frequent problems were seen in the introduction section, the sampling method of study, the exclusive criteria, analysis, and the limitations of the study.
CONCLUSION: Major deficiencies in the articles on addiction in Iranian journals were seen. It can be prevented by including important items in the checklists published by some of the journals for reviewers.

Introduction

The aim of science is to provide description, explanation, and prediction regarding a certain phenomenon. Generally, management of knowledge includes three stages: production of knowledge, publishing findings, and exploiting the results. If the produced knowledge is not published, evaluated, and criticized, it cannot be considered reliable or employed as a scientific tool or strategy.1 Writing good scientific articles is a skill. Many researchers seek to acquire this skill for the purpose of successfully disseminating their ideas to the scientific community.2 Unfortunately, articles and research reports presented in the scientific medical and psychological journals or in seminars and congresses are selected without a proper critical appraisal in many times, and this is a major problem since in many of these scientific journals, opinions of the experts are neglected. Perhaps, an important reason is the absence of trained personnel for criticism and/or lack of desire for criticism.2,3 Many physicians find themselves with less and less time to read what their coworkers have written, in spite of the increasing number of scientific publications. However, choosing, reading, and critical evaluation of publications are essential to keep up with the new results in one's field. This is also needed by the principles of evidence-based medicine.3,4 In addition to the content of a published study report, a comprehensive understanding of the statistical methodology is needed to interpret and evaluate as well. Unfortunately, many terms are used incorrectly in scientific articles. For instance, the concept of "significance" has been overused, since significant (or positive) findings are easier to get published.5,6

The main structure of scientific publications is always the same basically. The title, abstract, and keywords are presented before the main manuscript. The main frame of the article is separated into introduction, methods, results, and discussion (IMRAD), as well as conclusion and references at the end.7

Abstract: Abstract is a part summing up the aim and scope of the article. Studying the abstract presents, a general idea about its scientific quality rather reading the article from the beginning to the end. This advances the reading and helps to throw away improper studies initially, without spending too much time for them. In critical appraisal, the first stage is to search for an explanation of the problem addressed by the study: does a well-defined study question exist? This allows you to decide whether the research design is appropriately created to respond to the main question. The last sentence of introduction often represents the research question; however, for more details, you may need to look in the methods section. Another important issue is the explanation of the interventions. If you understand that the study is likely to include valid findings, you are ready to refer to the results section of the paper. If you think the study has serious errors, since a low-quality study cannot provide the reliable results, then you can skip the results. In the case of strong results in the abstract, it is quite important to read the provided tables before the text. Before leaving the abstract section, the ultimate question exists: does the article involve all the relevant outcomes, such as side effects?7 The last part of the abstract is conclusion finalizing the authors’ opinion through summing up their inference or main viewpoints.

Introduction: Introduction aims to explain the reader the investigation subject. The latest studies performed to address the subject should be provided with the reference to the recent literature. In this part, the essential of the study should also be stated obviously. The results of the studies mentioned in the review of the literature should be provided completely, reciting multiple results. Inaccurate phrases should be avoided such as "inconsistent findings", "somewhat better", and so on. In general, the text should indicate that the writer has fully read the cited articles. In case of uncertainty, it is suggested to the reader to consult these publications themselves. An optimal article backs up its central reports with references to the literature. Preferably, this section should proceed from general to more details. In the introduction, the questions which the study is intended to answer and the appropriate reason for the chosen design for this end are explained completely.7

Methods: This section tolerates a certain similarity to a cookbook. The description of the processes should be given to the reader as the "recipes" being followed to review the study. The essential data permitting the appraisal of the study validity can be found in this section.7 The methods section can be separated into subsections with their own headings; for instance, the investigation and intervention techniques and statistical methods can be explained separately. In the methods section, all the stages of planning, the composition of the study sample (e.g., patients, animals, and cell lines), the execution of the study, and the statistical methods should be described.7

Results: The findings should be presented clearly and objectively in this section, i.e., without interpretation. The explanation of the findings relates to the ensuing discussion. The results section should state directly the objectives of the study and be provided in a well-structured, readily understandable, and consistent manner. The results should first be formulated descriptively, affirming statistical parameters such as case numbers, mean values, measures of variation, and confidence intervals (CIs). A comprehensive description of the study population should be included in this section.7

Discussion: Owing to the presentation of the results, one has to select the more informative and less space-consuming form. Any description of data or measurement methods should be prevented in this section. In this part, the writer should explain his/her results truthfully and openly. Irrespective to the study type, two main goals essentially exist to be achieved: “comparison of the results with the other researches” and “critical analysis of the research limitations”.7

Rezayi Qale et al. have evaluated the output of medical research in Iran in terms of number of published articles in international journals.8 Moreover, in another article by Sheibaninia et al., the accuracy of writing in published dentistry articles in scientific researching journals has been studied.9 In terms of methodology and statistical expertise, international journals are not significantly different from local journals with regard to the design of studies and the accuracy of the data analysis. Furthermore, many of the papers published in international journals have major weaknesses in statistical methodology.10-13

So far, few researches have been carried out to evaluate the quality of articles in scientific and professional journals. Therefore, this article aims to evaluate the papers which address the issue of addiction in 8 Iranian journals.

Methods

Our descriptive study was performed in fall and winter 2016. Journals were identified searching the electronic databases, including PubMed, Medline, Scopus, Google Scholar, and national Persian databases of SID and Magiran. All articles on addiction published in the time span of 2010-2015 in 8 psychiatry, psychology, and addiction journals were included in this study. These journals included “Journal of Behavioral Sciences”, “Iranian journal of psychiatry and clinical psychology”, “Journal of Fundamentals of Mental Health”, and “Journal of research on addiction” in Persian, and “Health and Addiction Journal”, “International Journal of Psychiatry and Behavioral Sciences”, “Journal of Research in Behavioral Sciences”, and “International Journal of Psychological Studies” in English. These journals were evaluated based on research methodology, scientific writing, and evidence-based medicine references. The research design, the number of authors, and references were recorded for each paper. Subsequently, after covering the name of the authors and coding (blinded evaluation), all the papers were evaluated. We used Valaei et al.’s questionnaire,14 The questionnaire validity has been investigated by consulting experts in methodology and scientific writing. Various parts of the articles including title (2 items), abstract (11 items), introduction (6 items), methods (19 items), results (7 items), discussion (6 items), references (3 items), and general items were assessed by a total of 63 questions. The main elements assessed through these questions were whether the time and scenery of the research, research methodology, statistical tests, P-value for differences, clear answer for the main question, and a proper conclusion were provided in an article. Items covered in the introduction part were as follows: clear main aims, the location of the study, and logical sequence of the statements. In evaluating the methods part, it was revealed that whether or not the research design was emphasized and the time of the study, statistical tests, clear description of the main outcomes, and validity and reliability of measurements were all provided in the main text of the article. Another focus of the present study was to discover whether the applied examinations were proper and whether a pilot study was performed. The description of a number of studied cases, inclusive/exclusive criteria, randomization, and blindness of any type applicable were evaluated. In the questionnaire, the side effects of the intervention and patient satisfaction in clinical trials were also included. The items in the questionnaire were judged as proper, non-proper, and not applicable. Moreover, the number of authors in each article was calculated. The questions designed for the results section aimed to indicate whether or not the main consequence was stated clearly, all results were provided in one form and in an appropriate way, tables/figures were simple and accurate, and any of the explanations or conclusions existing in the results section was considered as inappropriate. Specific statistics should be mentioned along with the P-value. The main point in the discussion section was to avoid repetition of the introduction or results. Stating limitations of the study and posing new questions for future studies were evaluated as necessary. The discussion section should involve the clinical validation of the results and prove the external validity in case. In the reference part, employing papers in Persian was an advantageous point; however, references not used adequately were considered unsuitable. The style of addressing the references was also analyzed. SPSS software (version 17, SPSS Inc., Chicago, IL, USA) was used for a descriptive analysis.

Results

In the studied journals, 264 papers were published which included 5 (1.9%) editorial papers, 46 (17.4%) cross-sectional studies, 7 (2.7%) case reports, 2 (0.8%) brief reports, 13 (4.9%) review articles, 31 (11.7%) descriptive studies, 49 (18.6%) experimental studies, 1 (0.4%) cohort study, 5 (1.9%) qualitative studies, 59 (22.3%) case-control studies, and 46 (17.4%) correlational studies. Frequency and percentage of subject of these articles was as follows: 50 (18.9%) ones were prevalence, 99 (37.5%) were risk factors, 45 (17.0%) were prevention, 27 (10.2%) were treatment, 24 (9.1%) were relapse prevention, and 19 (7.2%) were other subjects (Table 1). The number of writers was calculated 1 to 7 with the mean of 3.5. The most frequent problem in different sections was that the design of the research could not be grasped by reading the title (Table 2). In the abstract part, time of study and statistical tests were missed in most articles. The most frequent problem in the introduction section was non-mentioning the study place in 68.6% of them (Table 2). In 64% and 62% of the papers a comprehensive explanation of sampling of study and exclusive criteria was not presented in the methods part, respectively (Table 3). In 68% of the papers, results section included analysis and interpretation. Limitations of the study and suggesting new questions and comments were not mentioned in the discussion part in 54% and 53% of the papers, respectively (Table 3). Moreover, 27.5% of references were Persian references and 63.5% of references were inappropriate. Furthermore, three options consisting of ethics, conflict of interests, and fund of projects were mentioned in 39%, 34%, and 20 % of articles, respectively.

Table 1

The subject of articles of the research reports on addiction in Iranian journals

Subject of research	n (%)
Prevalence	50 (18.9)
Risk factors	99 (37.5)
Prevention	45 (17.0)
Treatment	27 (10.2)
Relapse prevention	24 (9.1)
Others	19 (7.2)
Total	264 (100)

Table 2

Original papers appraisal of “title, abstract, and introduction” parts of the research reports on addiction in Iranian journals

Items of title	Not appropriate [n (%)]	Not applicable [n (%)]	Items of introduction	Not appropriate [n (%)]	Not applicable [n (%)]
Indicating research methodology	97 (36.7)	3 (1.1)	Beginning with problem	61 (23.1)	3 (1.1)
Place and time in the title of descriptive researches	125 (47.3)	116 (43.9)	Presenting a review of literature	92 (34.8)	3 (1.1)
Items of abstract	Not appropriate [n (%)]	Not applicable [n (%)]	Explaining the significance of study	110 (41.7)	5 (1.9)
Structuring of abstract	12 (4.5)	5 (1.9)	Presenting the main objectives clearly	21 (8.0)	4 (1.5)
The aim of research	11 (4.2)	7 (2.7)	Place of study	181 (68.6)	13 (4.9)
Time of research	167 (63.3)	15 (5.7)	Logical sequence of research statement	129 (48.9)	33 (12.5)
Place of research	101 (38.3)	15 (5.7)	Items of others	Not appropriate [n (%)]	Not applicable [n (%)]
Name of methodology	68 (25.8)	8 (3.0)	Conflict of interests	173 (65.5)	5 (1.9)
Presenting statistical test	154 (58.3)	11 (4.2)	Acknowledgments	166 (62.9)	5 (1.9)
Presenting P-value	142 (53.8)	19 (7.2)	Financial resources	211 (79.9)	5 (1.9)
Mentioning the results	34 (12.9)	11 (4.2)	Resulting thesis or approved project	169 (64.0)	21 (8.0)
Mentioning the conclusion	16 (6.1)	8 (3.0)	Authors’ contribution	236 (89.4)	7 (2.7)
Appropriate conclusion	43 (16.3)	8 (3.0)	Statistics or epidemiology specialist in authors	231 (87.5)	10 (3.8)
Presenting clear answer	42 (15.9)	5 (1.9)	-	-	-

Table 3

Original papers appraisal of “methods, results, and discussion” parts of the research reports on addiction in Iranian journals

Items of methods	Not appropriate [n (%)]	Not applicable [n (%)]	Items of results	Not appropriate [n (%)]	Not applicable [n (%)]
Presenting the research design	47 (17.8)	11 (4.2)	Considering the main result	21 (8.0)	43 (16.3)
Presenting time of study	105 (39.8)	14 (5.3)	One result, appropriate way	44 (16.7)	24 (9.1)
Presenting place of study	36 (13.6)	17 (6.4)	Appropriate table	26 (9.8)	36 (13.6)
Presenting statistical test	80 (30.3)	24 (9.1)	The accuracy of title of tables and charts	23 (8.7)	55 (20.8)
Presenting the main outcome	69 (26.1)	68 (25.8)	Correct figures	6 (2.3)	180 (68.2)
Presenting explanation of measurements	27 (10.2)	27 (10.2)	Analysis and interpretation in result section	59 (22.3)	26 (9.8)
Presenting explanation of the sampling of study	170 (64.4)	30 (11.4)	Specific statistics, P-value	23 (8.7)	31 (11.7)
Reliability	54 (20.5)	56 (21.2)	Items of discussion	Not appropriate [n (%)]	Not applicable [n (%)]
Validity	95 (36.0)	54 (20.5)	Repetition of the introduction	189 (71.6)	18 (6.8)
Consort diagram	81 (30.7)	175 (66.3)	Repetition of the results	143 (54.2)	14 (5.3)
Inclusive criteria	119 (45.1)	28 (10.6)	Limitation of study	111 (42.0)	11 (4.2)
Exclusive criteria	164 (62.1)	29 (11.0)	Disclosing new questions and comments	109 (41.3)	14 (5.3)
Randomization	135 (51.1)	54 (20.5)	Conclusion	150 (56.8)	7 (2.7)
Blindness	76 (28.8)	174 (65.9)	External validity (generalizability of findings)	159 (60.2)	17 (6.4)
Duration of follow-up	53 (20.1)	162 (61.4)	Writing style of “references”‌	n (%)
Side effects	73 (27.7)	169 (64.0)	Vancouver	219 (83.0)
Patients’ satisfaction	121 (45.8)	56 (21.2)	Harvard	45 (17.0)
Doing a pilot study	125 (47.3)	68 (25.8)	Total	264 (100)

Discussion

In the present study, the number of authors of published articles was 1 to 7 persons with a mean of 3.5. The highest number of authors of articles was 2, 3, and 4, and articles with 3 authors were more frequent than others. Type of study was not mentioned in the abstract section in 26% of articles. Most of the papers were descriptive, experimental, and laboratory studies, while the lowest percentage of articles included quasi-experimental, qualitative, cohort, and self-assessment clinical studies. In a study, Persian to English references ratio was 2.8 to 19.3,15 but in our study 27.5% of references were Persian and the others were English. Moreover, most of the articles were descriptive and the least was cohort. In a study conducted by Rezaeian et al., it was indicated that the most published articles in Journal of Rafsanjan University of Medical Sciences, Rafsanjan, Iran which is a general medicine journal, were descriptive, and experimental laboratory studies were not frequent.16

Further, in this study, in 90 articles (47.6%) the type of study had not been mentioned and mean of authors’ number was 3.4.16

Also in the study conducted by Rezaeian et al., descriptive studies were higher than other studies.16 Undoubtedly, problems such as time financial support, facilities, and expertise limitations have a key role in the abundance of descriptive researches. In experimental studies, the nonexistence of financial support is a main problem. In a study conducted by Valaei et al., 182 articles in the Journal of Mazandaran University of Medical Sciences, published during the years 2002-2006, were evaluated; the results showed that 63% of the introductions were unsuitable in writing. In addition, in 56.4% of them, the time, place, and purpose of the study were not mentioned. Moreover, only 20% of articles had a methodologist in authors.14 In our study, the time and place in the methods section were not mentioned in 40% and 14% of the articles, respectively.

In a study performed by Sheibaninia et al.,9 it was indicated that the purpose of study in 90% of articles published in Iranian scientific journals of dentistry in 2006 was not expressed correctly, and also a proper justification for publishing was not provided in 72% of the articles. Moreover, most of the articles did not involve any reference to Persian sources. They have reported that the main question was not solved in 68% of articles, whereas in our study, in 16% of articles, the main question was not solved. Reliability of the employed questionnaires was not mentioned in 96% of articles, but this amount was 20.5% in our study. Explanation of sampling was not mentioned in 93% of the studied journals in comparison to 64% in our study. In 87% of cases, reporting limitations of the study was missed in comparison to 42% in our study. Regarding their 52 studied variables, the main defects were found to be the lack of instrument validity report, the lack of feasibility of reporting sample size, the lack of reported weaknesses in the discussion, inaccuracies in title and name of rows and columns of tables, the absence of the aim of study in the introduction, the length of the introduction, and incorrect statistical tests, for example, 61% of statistical tests were incorrect in these articles.9

Heidari et al. reported that 85.5% of articles in scientific journals of nursing-midwifery in Tehran, Iran, did not state the objective, place, and time appropriately in the introduction. Also, 85% of them did not present a logical reason for research. Sampling in 78.7% and blindness in 74% of the papers were not mentioned. Lack of reporting the follow-up period was 6%. 87.5% of the papers contained errors in rows and columns of the tables. Inaccurate statistical tests were used in 53% of them, a result that is similar to the findings of our study. Regarding the discussion part, incorrect writing and weaknesses were observed in 95.2% and 85.5% of the papers, respectively. Lack of suitable analysis in 74.7% and lack of correct conclusion in 41.0% of them were detected. Writing of references was correct in 74.7% of papers;17 though, these problems were seen in our study. The mean number of Persian articles was 4.2 ± 4.1 and the mean number of new references was 5.6 ± 4.2. Moreover, the weakest part was introduction in 76.5% of articles. Weakness in discussion and results were reported in 53.7% and 56.9% of articles, respectively, that are comparable with our study.17 In a study, the mean number of authors was 2.90 ± 0.84 that is compatible with our study. The most frequent type of study was experimental-in vitro study (46%), followed by descriptive (33%), clinical trial (10%), and historical cohort (6%). The frequency of instrument validity report absence was 96%, that in our study, validity and reliability were not mentioned in 36% and 20.5% of articles, respectively. No explanation of sample size in 93% and lack of limitation and important findings report in 87% of articles were observed,9 whereas in present study explanation of the sampling of study was not mentioned in 64% of articles.

In ISI medical journals, the number of defects in the tables and graphs was 7.9%,18 that is compatible with our study. In addition, the accuracy of the title of tables and charts was 92% in the articles of our study.

In a study by Jawaid et al., 9.7% of tables and figures of the articles of Pakistan Journal of Medical Sciences had some defects, that is compatible with our study with 9.8% defect in this section. Also, writing references were correct in 75% of them.18 Valaei et al. reported that in articles from 2001-2005, 2.2% of articles had not reported P-value that in our study it was 39%. In discussion part, repeating the findings and repeating the introduction was seen in 7.2% and 21% of the articles, respectively; that in our study it was 72% and 54%, respectively. And in 0.6% of the articles, the findings were repeated both in text and tables that in our study it was seen in 17% of the articles. There was a statistics or epidemiology specialist in authors in 20% of the articles that in our study it was seen in 13% of articles. In addition, in 37% of the articles, the research method in abstract was not mentioned that in our study it was 26%.14

A study by Kanter and Taylor which was conducted to evaluate the accuracy of statistical methods in published articles in Transfusion Journal reported that 80% of the tests were incorrect,19 Kurichi and Sonnad showed that 27% of the published articles in surgery journals employed incorrect statistical methods.20

Numerous deficits in articles on addiction in Iranian journals highlight the importance of the fact that the researchers and reviewers of the journals need more training for research methodology and scientific writing. This fact has also been highlighted by other authors.18-21 If we pay attention that many other rejected manuscripts with more problems were not studied, the necessity of training of researchers and authors will be more highlighted. Lack of assessment of weightage of each deficiency is another limitation. As mentioned by Jawaid et al., this is an inclusive deficit of other studies, too.18

Conclusion

Major deficiencies in the articles on addiction in Iranian journals were as following: not mentioning of place and time of the study in the title of descriptive researches, not mentioning of statistical tests in abstract section, and not mentioning of place of study in introduction section in many of the studies. In methods section, not explanation of the sampling of the study and exclusive criteria, and in discussion section, repetition of the introduction and results were major deficiencies. Conclusion and limitation section was not mentioned in many of the articles. Moreover, statistician or epidemiologist was not found in the list of authors in several articles. Also, authors’ contribution and financial resources were not mentioned in many of the articles. These can be prevented by including these items in the check lists published by some of the journals for reviewers.