A bibliometric analysis of bipolar affective disorders using density-equalizing mapping and output benchmarking.

BACKGROUND: Bipolar affective disorder (BaD) has a great impact on health systems worldwide. Although bibliometric studies have been done on this subject, these studies did not do an analysis of the contents of papers, the cooperation between countries, or of the names currently used to describe the condition. Furthermore, the number of publications since the last bibliometric study has doubled. AIM: This study was to examine the recent developments in the field, explore main topics/subject areas of the top 10 authors in this research field, and to compare diagnosis-defined data between International Classification of Diseases, 10(th) edition (ICD-10) and Diagnostic and Statistical Manual of Mental Disorders, fourth edition (DSM-IV).
MATERIALS AND METHODS: Using distinct search terms, the Web of Science database developed by the Thompson Reuters Institute of Scientific Information was scanned for relevant items published between 1900 and 2008. Results were analyzed using scientometric methods and density-equalizing calculations.
RESULTS: We found an important increase of publications on the subject over the last decade. Most published studies came from North America and Europe, while the countries cooperating with each other were comparable to other areas of medical research. Although there has been an increase in publications on BaD (m=3.3 publications per year in the last decade), the number of works using the term bipolar disorder (BD) was considerably higher (m=141.8 publications per year in the last decade). We found that the subject areas, genetics and pharmacology were focuses of research for the 10 most prolific authors, all of whom where psychiatrists.
CONCLUSION: Research interest in BaD is rising. Reflecting the two main disease classification systems, DSM and ICD, both terms BD and BaD are used in research, with a preponderance of the former. The research of the most prolific authors engages genetic and pharmacological questions.

INTRODUCTION

Bipolar affective disorder (BaD) is a multifactorial disease with a wide range of clinical manifestations.[1] The same condition is described under slightly different clinical classification systems:[2] The International Classification of Diseases (ICD)-10 refers it as “bipolar affective disorder”[3] and the Diagnostic and Statistical manual for Mental Disorders (DSM)-IV-TR as “bipolar disorder” (BD).[4] According to the ICD-10, BaD is characterized by two or more episodes in which the patient's mood and activity levels are significantly disturbed.[3] This disturbance can on some occasions consist of an elevation of mood and increased energy and activity levels (hypomania or mania) and on others of a lowering of mood and decreased energy and activity (depression).[3] The DSM-IV-TR classifies the condition as BD, according to the course of the disorder as type I (with severe manic or mixed episodes and a major depressive episode, usually requiring hospital admission) and type II (where there have been no manic or mixed episodes, but one hypomanic episode and a major depressive episode).[4]

Although manic and depressive symptoms were described by Hippocrates and Aristotle, the existence of a condition with both manic and depressive features was first documented by Aretaeus of Cappadocia, who lived in Turkey from 30 to 150 AD.[5] He hypothesized that both “melancholy” and “mania” had a common origin in “black bile.” This condition has had different names in the last 200 years. In 1854, Jules Falret described the condition as “folie circulaire” (circular insanity), and Emil Kreaplin described it as “manic-depressive psychosis” in the early 1900s.[5] This was the most widely accepted term in the 1930s and appeared in the ICD for the first time in the ICD-5 in 1938.[6] In the ICD-6 of 1948[7] and the ICD-7 of 1955,[8] it was changed to “manic-depressive reaction,” while in the ICD-8, the term “manic depressive psychosis” was once again used in 1965.[9] The name “bipolar affective disorder” has been used since then in the ICD-9 and -10.[210]

The first two editions of the classification of mental disorders according to the DSM-I in 1952 and DSM-II in 1968 used the same term as the ICD-7 and the ICD-8,[11] using the term “manic-depressive illness (manic-depressive psychosis).”[12] From the DSM-III of 1980 onward, the condition was referred to as “bipolar disorder.”[2] This includes the DSM-III-R of 1987, the DSM-IV of 1994, and the DSM-IV-TR of 2000. The latter two versions also distinguish between bipolar disorder types I and II.

BaD affects patients worldwide, with a mean age of onset of 21 years.[13] While earlier studies have projected the prevalence of BaD in a range of 1–2%,[14] these numbers might even be higher because many studies do not include milder forms of the disease.[15]

The disorder has also gained medical attention as a high suicide rate and the potential for sexual abuse (females) and aggression (males) has been observed.[16] A recent study in 2009 also found high premature mortality from general medical conditions like cardiovascular, respiratory, cerebrovascular, and endocrine disorders in patients with BaD.[16]

Different treatment strategies have been tested including psychopharmacology, psychotherapy, psycho-education, and even electroconvulsive therapy.[17-19] From a psychopharmacological perspective, the effect of lithium salt on mood changes was first published in 1949 by Frederick Joseph Cadein.[20] Lithium as treatment for BaD has proven its efficacy through various double-blinded trials during the 1970s and is regarded as the gold standard for treatment.[2122] Since then, sodium valproate and carbamazepine have become established as other options for the long-term treatment of this condition. Another more recent approach is the use of atypical antipsychotic drugs like olanzapine, risperidone, and quetiapine.[23] To standardize pharmacotherapy, various guidelines have been published; examples are the guidelines of the British Association for Psychopharmacology in 2003[19] and more recently the NICE guidelines in 2006.[24]

The above-mentioned historical developments have resulted in a vast number of publications. Since the last two bibliometric studies of this condition in 2002 and 2004,[2526] the number of publications has doubled. Also, a qualitative analysis of the publications on BaD and BD has not been performed yet.

Therefore, the aim of the current study was to evaluate the scientific output for BaD and BD from 1900 to 2008 by bibliometric/scientometric means. Both terms were analyzed chronologically and in terms of qualitative bibliometric parameters such as average citation rates (acrs). Further analyses of the origin of the publications, subject areas, and cooperation between countries in this field were carried out.

MATERIALS AND METHODS

Data source

Data for this study were retrieved from the database Web of Science from the Thomson Reuters Institute for Scientific Information (WoS).[2728]

Search strategies

The terms “bipolar affective disorder*” and “bipolar disorder*” were used to include all relevant publications. The asterisk was included in order to encompass all possible endings. Additional searches were conducted using the Boolean Operators and NOT

“bipolar affective disorder*” AND “bipolar disorder” (for all publications that used both terms)

“bipolar affective disorder*” NOT “bipolar disorder*” (for publications that used the term BaD, but not BD)

“bipolar disorder*” NOT “bipolar affective disorder*” (for publications that used the term BD but not BaD).

Time span

The time span analyzed was from 1900 to 2008. 2009 could not be included since data acquisition has not been completed during the study period.

Citation quantities

The “citation report” function was used in order to analyze the published items. The acr is based on how often publications are cited on a specific topic, divided by the number of publications. This result indicates research quality of a specific journal or author. The following citation databases were used:

Science Citation Index Expanded (SCI-EXPANDED) – 1900 to present

Social Sciences Citation index (SSCI) – 1956 to present

Arts & Humanities Citation Index (A&HCI) – 1975 to present.

Density-equalizing mapping

In this method, specific calculations based on Gastner and Newman's algorithm[29] were used in order to produce a map of the world in which countries were sized according to specific variables, in this case the number of publications on BaD (and BD) and the acrs of these countries.

RESULTS

Total number of publications

Between 1900 and 2008, a total of 18,831 published items were found. In these, the term BaD appeared in 5445 published items, while all articles used the word BD. 13,386 publications utilized BD but not BaD, whereas all the publications that contained BaD also used BD.

From 1969 to 1978, publications were published at a gradient of m = 0.5 publications per year for BaD and m = 0.8 publications per year for BD. This increased to a gradient of m = 1.9 for BaD from 1979 to 1988 and m = 3.7 for BD. The gradient rose from 1989 to 1999 to m = 27.3 for BaD and m = 59.6 for BD. The last decade (1999-2008) witnessed a decrease in publications using BaD to m = 3.3, while publications with BD continued to increase to m = 141.8 publications per year [Figure 1].

Figure 1

Total number of publications relating to the terms “bipolar affective disorder” (BaD) and “bipolar disorder” (BD) in the Web of Science database from 1900 to 2008

Analysis of origin

Publications on BaD originated from 81 and on BD from 102 countries. The special distribution of publications on BaD and BD was comparable. Of the 5445 publications on BaD, 2386 (43.8%) were published by the USA, 853 (15.7%) by the UK, 610 (11.2%) by Germany, 382 (7%) by Canada, and 315 (5.8%) by Italy. The top five countries of the 18,831 published items containing BD were the USA with 9443 (50.2%), the UK with 2216 (11.2%), Germany with 1364 (7.2%), Canada with 1350 (7.2%), and Italy with 1044 (5.5%) [Figures 2 and 3, Table 1].

Figure 2

Density-equalizing map illustrating the number of BaD publications in each country. The area of each country was scaled in proportion to its total number of publications on BaD from 1900 to 2008

Figure 3

Density-equalizing map illustrating the number of BD publications in each country. The area of each country was scaled in proportion to its total number of publications on BD from 1900 to 2008

Table 1

A comparison of the top 10 publishing countries on BaD and BD from 1900 to 2008

Average citation rates of countries

To evaluate the quality of published items, we analyzed the acrs of countries for the search term BD. Due to a bias that results when countries that publish few papers that are cited a lot, the acr was only analyzed for countries that published at least 30 papers. We found that Costa Rica had the highest acr (42.9), followed by Finland (29.5), New Zeeland (25.3), and the USA (25.2). Of these countries, Costa Rica published the least papers (35), followed by New Zeeland (131), Finland (135), and the USA (9406). A density-equalizing map in which the area of countries with more than 30 publications was scaled in proportion to their acr depicts this data [Figure 4].

Figure 4

Density-equalizing map illustrating the average citation rate of the countries that published at least 30 papers on BD from 1900 to 2008. The area of each country was scaled in proportion to its average citation rate on BD publications

Multinational cooperation

The analysis of multinational cooperation was also limited to the BD publications. We found that the USA published the largest number of papers with other countries (1929 papers), followed by the UK (737 papers) and Germany (564 papers). Figure 5 presents a graphic presentation of the cooperation between these countries.

Figure 5

A comparison of countries with collaborative publications on BD from 1900 to 2008. Format: Country (number of publications/number of publications in cooperation with other countries)

Analysis of the research content

In order to recognize important subject areas in the research output from 1900 to 2008, we assessed the publications of the 10 most prolific authors for BD. We found that these authors published 1988 papers. Of note was the high number of papers on pharmacology (644, 32%) and genetics (209, 11%). Other recurring areas were puerperal psychosis, bipolar type II, rapid cycling, and neurological studies [Table 2]. We also analyzed the academic background of these 10 authors and found that all 10 were consultant psychiatrists.

Table 2

Comparison of the main subjects of the top 10 authors who published papers on “bipolar disorder” from 1900 to 2008

DISCUSSION

This study was carried out to examine the recent developments in the field of BaD, explore main topics of the 10 most prolific authors, and to compare diagnosis defined data between ICD-10 and DSM-IV. Countries of origin were identified, as well as their acr and the extent of multinational cooperation.

Recent studies have found a great interest in BD with an increasing number of publications on the topic.[2526] Our data established a continuation of this situation with a doubling of published items as compared to the data of the former studies.[2526] This trend has been demonstrated in other medical fields as well and might have various reasons.[30-32] Simplified international communication structures through the internet and enhanced political interest in cooperation might all have contributed to the increase.[33] A change in the search algorithm introduced by the ISI database system in 1990 has also enhanced the accessibility of publications.[34] The increase in the current field has been found to be not as pronounced as in other psychiatric fields, like schizophrenia.[25]

The number and distribution of publishing countries was comparable to the one found in other medical fields (heart disease, scoliosis, psoriasis, and rheumatoid arthritis),[30-323536] with the most prolific countries situated in North America and Europe. All top five countries were from these regions, with the USA being the most prolific. When comparing the available prevalence rates of BD for these countries (USA 3.9, UK 0.5, Germany 6.5, Canada 1.1, and Italy 2.6),[3738] it becomes clear that there is a discrepancy between the number of publications and the prevalence rates of these countries. This discrepancy was also found in a similar study on Leishmaniasis.[39] This discrepancy could be attributed to a lower relative interest in BD research (compared to other psychiatric research) for Germany, Canada, and Italy, as was found by Lopez-Munoz et al.[26]

We found that the US cooperated most with other countries. Cooperation might be a way to enhance the research output. In former studies done by our department, we found that countries with important scientific output tend to cooperate more with other countries.[40-42] We found that all the 10 most prolific publishing countries cooperated a lot with other countries. The two countries that cooperate a lot with each other in this and other studies[30-3239] are the United States and the UK. This could be attributed to the fact that most of the BD-related pharmaceutical companies are situated in these two countries.[26]

The acr is a further indicator of the visibility of the countries’ research in the field. Besides high-quality research, there are other reasons for a country attaining high acr. A certain quantity bias results in countries with lesser publications, but achieving higher acr. To partially circumvent this problem, a limit of a minimum number of publications was used in the study. Also, it has been found that the rate of self-citation might be different. In a recent bibliometric study on psoriasis, Finland was found to have a disproportional high self-citation rate.[35] In our study, Finland was the country with the second highest acr (Costa Rica had the highest), followed by New Zealand and the USA. The high acr of Costa Rica could be attributed to the small number of publications (35), as the other three countries published at least 3 times as many papers. The acr values of this study are similar to those of bibliometric studies in other medical fields.[39]

Evaluating the most frequent subjects in the field, pharmacology and genetics were two centers of attention of the 10 most prolific authors. The focus on pharmacology could be due to the emergence of anti-psychotic drugs, like olanzapine, risperidone, and quetiapine, as new treatment options for BD.[23] The use of these drugs has also been recommended in combination with the established drugs lithium, sodium valproate, and carbamazepine, that have been in use since the 1970s.[23] This finding corresponds with older bibliometric data on BD.[26] Another subject that is a center of attention for these authors is genetics. The genetic subject was first mentioned in a study by Falret in 1854, which demonstrated that BD runs in families.[43] At the end of the 21st century, genetic research gained momentum with the advance of modern molecular techniques and the possibility of identifying genes that contribute to patients developing BD.[44] The underlying hope of these efforts is the expectation that advances in the understanding of the genetic make-up will pave the way for enhanced diagnostic accuracy and improved treatments.[44]

We found that the gradient of increase in the number of publications that used the term BD was considerably higher than the gradient for BaD. Both terms are part of different disease classification systems (ICD and DSM). The spread of both terms could be dependent on the prevalence of the corresponding disease classification system in specific countries. We did not find evidence for this assumption, as the top five publishing countries for BaD and BD were the same, with slight variations in the percentages. A second possible reason could be that the DSM system is currently being preferred for research to the ICD system. Maser et al. found in an international survey in 1991 that the DSM system (in this case DSM-III) was used more widely than the ICD (in this case ICD-9).[45] Another international survey done 11 years later (2002), using the DSM-IV and the ICD-10, found a different result. In this survey, the ICD-10 was used more than the DSM-IV.[46] The reasons given in the latter study were that psychiatrists preferred the ICD-10 for clinical diagnosis and training, but preferred the DSM-IV for research.[46] The fact that research is still being done using the ICD term (BaD) is confirmed by the fact that there is still an increase in the number of papers being published, although this increase is considerably less than that for DSM term (BD) over the last decade. Our study would therefore confirm the finding that the DSM-IV is currently being used more for research than the ICD-10.