Author, study, and publication metrics of brief reports and full-length research articles in ophthalmology journals

The brief report is a type of original research article that carries limitations on the number of words, tables, or figures allowed.¹ Brief reports facilitate the communication of preliminary findings, may present novel findings, and are often published in connection with full length articles; in this pairing, the initial article may be either the brief report or the full-length article. This cross-sectional study investigated in the ophthalmic literature the differences between (a) brief reports and full-length articles and (b) paired and unpaired brief reports.

We defined a brief report as an original research article type limited to 2000 words or less and reviewed editorial policies to determine whether journals accepted brief reports. We included the 10 ophthalmology journals with the highest impact factors that (a) accepted unsolicited brief reports and (b) did not require article processing charges for all submissions, as journals with fees often have different editorial standards (Supplementary Material 1).

A literature search strategy was developed in accordance with Preferred Reporting Items and Meta-Analyses (PRISMA) guidelines (Supplementary Material 2).² Two investigators (CK; JCL) independently conducted title/abstract screening, full-text review, and data extraction in Covidence. Disagreements were resolved by the senior investigator (PBG).

We recorded brief reports or full-length articles published in 2016 that covered a topic related to the leading causes of blindness and vision impairment in the United States (age-related macular degeneration, cataract, diabetic retinopathy, and glaucoma).³

We determined whether the brief report had a paired full-length article within 5 years by searching PubMed and Google Scholar. Matches had identical study question, design, eligibility criteria, conditions, and at least one author in common. If multiple full-length articles were linked to a single brief report, the first full-length article was recorded.

Characteristics of publications and first/corresponding authors, changes in declarations of interest, funding sources, time in months between publications, and outcome order were recorded. Substantial variation in sample size was defined as a difference greater than 10%.

After screening, 863 full-length articles and 122 brief reports were included (Supplementary Material 3). Relative to full-length articles, brief reports were more likely to use retrospective and cross-sectional designs, be conducted at a single centre, report statistically non-significant results, and lack funding (Table 1). Brief reports were also more likely to have fewer authors and citations and have first and corresponding authors with longer publication records.

TABLE 1.

Comparison of brief reports and full-length articles

	Brief reports, n = 122	Full-length articles, n = 863	Statistic, p value
Author characteristics
Number of authors, mean (SD)	4.76 (2.46)	6.25 (2.92)	t = −5.37, p < 0.001
1–5, n (%)	90	382	χ2 = 37.29, p < 0.001
6+	32	481
First author gender, n (%)
Male	72	522	χ2 = 0.37, p = 0.541
Female	48	308
First author degrees, n (%)
No doctoral degree	5	61	χ2 = 7.69, p = 0.053
MD or MD-equivalent	73	382
MD/PhD	17	130
PhD only	18	172
First author publications, mean (SD)	54.12 (86.31)	70.35 (129.99)	t = −1.34, p = 0.181
0–10, n (%)	40	223	χ2 = 2.78, p = 0.249
11–70	55	415
71+	27	225
First author h-index, mean (SD)	12.04 (13.64)	15.64 (18.07)	t = −2.11, p = 0.035
0–5, n (%)	49	228	χ2 = 10.49, p = 0.005
6–20	53	433
21+	20	202
First author m-quotient, mean (SD)	1.39 (0.92)	1.56 (1.41)	t = 1.33, p = 0.183
0–0.5, n (%)	18	86	χ2 = 2.64, p = 0.267
0.51–1	33	255
1.01+	71	522
First author years of experience, mean (SD)	9.29 (7.58)	13.35 (68.04)	t = −0.66, p = 0.511
0–10, n (%)	77	497	χ2 = 4.05, p = 0.132
11–20	30	194
21+	15	172
First author institution, n (%)
Academic	112	818	χ2 = 6.04, p = 0.109
Government	4	7
Industry	2	11
Private Practice	4	27
First author country income, n (%)
High income	103	711	χ2 = 3.35, p = 0.187
Upper-middle income	9	104
Lower-middle income	10	48
Low income	0	0
Corresponding author gender, n (%)
Male	85	264	χ2 = 0.23, p = 0.628
Female	35	578
Corresponding author degrees, n (%)
No doctoral degree	7	27	χ2 = 5.77, p = 0.123
MD or MD-equivalent	65	391
MD/PhD	24	227
PhD	22	180
Corresponding author publications, mean (SD)	17.01 (14.83)	23.97 (20.86)	t = −3.56, p < 0.001
0–25, n (%)	54	236	χ2 = 12.38, p = 0.002
26–150	31	310
151+	0	1
Corresponding author institution
Academic	111	818	χ2 = 1.30, p = 0.729
Government	2	6
Industry	3	15
Private Practice	3	24
Corresponding author country income
High income	101	708	χ2 = 4.91, p = 0.086
Upper-middle	7	105
Lower-middle	10	50
Paper characteristics
Study topic
Cataract	33	271	χ2 = 1.59, p = 0.811
Glaucoma	34	236
Diabetic retinopathy	25	147
Age-related macular degeneration	20	131
Mixture	10	78
Study field
Basic science	17	190	χ2 = 6.82, p = 0.078
Clinical	94	580
Public health	11	79
Education	0	14
Study timing
Both	0	6	χ2 = 9.34, p = 0.009
Retrospective	50	247
Prospective	65	576
Study design
Case–control	18	130	χ2 = 12.98, p = 0.002
Clinical trial	44	444
Cross-sectional	54	252
Statistically significant results
No	52	159	χ2 = 37.19, p < 0.001
Yes	70	704
Sample size, mean (SD)	206.27 (124.97)	225.05 (126.65)	t = −1.51, p = 0.131
0–120	36	200	χ2 = 2.73, p = 0.255
121–300	49	321
301+	34	287
Sites
Single centre	98	597	χ2 = 5.80, p = 0.016
Multicentre	17	199
Open access
No	35	243	χ2 = 0.01, p = 0.903
Yes	87	620
Study funding
No	77	297	χ2 = 37.38, p < 0.001
Yes	45	566
Journal impact factor, mean (SD)	4.23 (1.23)	4.16 (1.24)	t = 0.59, p = 0.554
0–3.6	30	281	χ2 = 3.14, p = 0.076
3.7+	92	582
Citations, mean (SD)	13.31 (16.40)	25.52 (32.59)	t = −4.05, p < 0.001
0–10	75	252	χ2 = 51.46, p < 0.001
11–25	26	360
26+	20	251

Abbreviations: RCT, randomised controlled trial; SD, standard deviation.

In multiple logistic regression (Table 2), original research articles were more likely to be published as a brief report compared with a full-length article if they had statistically non-significant findings [OR = 0.27, 95% CI = (0.17–0.42)], absence of funding [OR = 0.36, 95% CI = (0.23–0.56)], lower first author h-index [OR = 0.60, 95% CI = (0.39–0.95)], and publication in a higher-impact journal [OR = 1.84, 95% CI = (1.13–2.98)].

TABLE 2.

Characteristics associated with brief report type in logistic regression

Author characteristics	Simple regression, OR (95% CI)	Multiple regression, OR (95% CI)
First author gender
Male	[Reference]	[Reference]
Female	0.89 (0.60–1.31)	0.87 (0.56–1.32)
First author h-index
0–5	[Reference]	[Reference]
6+	0.53 (0.36–0.79)	0.60 (0.39–0.95)
First author country income
LMIC	[Reference]	[Reference]
High income	1.16 (0.69–1.95)	0.93 (0.53–1.63)
Study field
Basic science	[Reference]	[Reference]
Clinical	1.81 (1.05–3.11)	0.96 (0.51–1.81)
Public health	1.56 (0.70–3.47)	0.86 (0.35–2.11)
Study design
Cross-sectional	[Reference]	*Dropped due to collinearity with field
Case–control	0.65 (0.36–1.15)
Clinical trial	0.46 (0.30–0.71)
Statistically significant results
No	[Reference]	[Reference]
Yes	0.30 (0.20–0.45)	0.27 (0.17–0.42)
Sample size
0–120	[Reference]	[Reference]
121–300	0.85 (0.53–1.35)	1.04 (0.63–1.71)
301+	0.66 (0.40–1.09)	0.79 (0.46–1.36)
Open access
No	[Reference]	[Reference]
Yes	0.97 (0.64–1.48)	1.32 (0.81–2.15)
Study funding
No	[Reference]	[Reference]
Yes	0.31 (0.21–0.45)	0.36 (0.23–0.56)
Journal impact factor
0–3.6	[Reference]	[Reference]
3.7+	1.48 (0.96–2.29)	1.84 (1.13–2.98)

Note: Statistically significant relationships are bolded.

Of 122 brief reports, 31 (25%) had paired full-length articles (Supplementary Material 4); 17 full-length articles had subsequent brief reports (14%) with a mean publication time of 27.9 months and 14 brief reports had subsequent full-length articles (11%) with a mean publication time of 29.6 months (Supplementary Material 5). Brief reports with paired full-length articles compared with unpaired brief reports were more likely to have higher number of authors (p = 0.043) and first authors with higher degrees (p = 0.0178) and h-index (p < 0.00001).

Our study found that brief reports were cited less often than full-length articles, confirming the findings of a previous microbiology study.⁴ Brief reports may have smaller scopes of research or focus more on hypothesis generation, which have been negatively associated with citation counts.

As brief reports relative to full-length articles are less likely to report statistically significant findings, they may help reduce publication bias. Accordingly, systematic reviewers should consider including brief reports in their eligibility criteria when executing their search strategies.

Like abstracts, brief reports often present preliminary data that warrants further investigation. We found that 11% of ophthalmic brief reports had a later full-length article, which is lower than the publication rates for conference abstracts in other medical specialties.⁵ This may be due to time constraints, competing commitments, lack of funding, or relocation of authors. Our findings suggest that larger research teams with more experienced authors, a prospective study design, and statistically significant findings may be more likely to follow a brief report with a full-length article.

Strengths of this study include a systematic approach to identifying articles for inclusion and a comprehensive analysis of ophthalmic brief reports and full-length articles. Study weaknesses include potentially limited generalizability and self-citations in Google Scholar.

In summary, this study suggests that in the ophthalmic literature, brief reports significantly differ from full-length articles in both publication and author characteristics. Brief reports can be a valuable resource to systematic reviewers andmay also play a role in reducing publication biases. Further investigation is necessary to better elucidate the underlying factors for the low rate of paired brief report-full-length articles.