|Year : 2019 | Volume
| Issue : 1 | Page : 55-62
Effectiveness of core stability exercises in patients with chronic non-specific low back pain: A review of randomized controlled trials
Fatima Kachalla Gujba1, Nicky Lambon2, Sokunbi Ganiyu3, Mamman Ali Masta1, Muhammad Ali Usman1
1 Department of Physiotherapy, University of Maiduguri, Maiduguri, Borno, Nigeria
2 Department of Physiotherapy, Coventry University, Coventry, United Kingdom
3 Department of Physiotherapy, Bayero University Kano, Kano, Nigeria
|Date of Web Publication||13-Sep-2019|
Prof. Sokunbi Ganiyu
Department of Medical Rehabilitation (Physiotherapy), University of Maiduguri, Maiduguri, Borno State, P.M.B 1069
Source of Support: None, Conflict of Interest: None
Individual studies have shown that core stability exercise (CSE) program is effective at improving pain and function in patients with chronic nonspecific low back pain (LBP) and the use of this intervention has become increasingly popular. However, there is still a further need for systematic review to infer the precise efficacy of CSEs for chronic nonspecific LBP (CNLBP). The aim of this study was to systematically review randomized controlled trials (RCTs) evaluating the effectiveness of CSEs in patients with CNLBP. A systematic review of RCTs was done using published articles. Recognized databases such as CINAHL, AMED, MEDLINE, PEDro, and The Cochrane Library were used to search for RCTs published between 2010 and 2015 in which pain and disability were evaluated as outcomes. Methodological quality was assessed using the PEDro scale. Six studies met the criteria for this review. The included studies randomised participants into two different exercise groups. All studies had PEDro scores of >5/10. Five out of the six studies showed benefits of CSEs over other intervention or control for pain and disability while the other study shows both CSEs and traditional trunk exercises to be effective. The result of this review supports the effectiveness of CSEs at improving pain and disability in patients with CNLBP. Further studies with robust methodology are warranted to assess the longterm effects of this intervention in patients with CNLBP.
Keywords: Chronic nonspecific low back pain, core stability exercises, effectiveness, low back pain, randomized controlled trials, systematic review
|How to cite this article:|
Gujba FK, Lambon N, Ganiyu S, Masta MA, Usman MA. Effectiveness of core stability exercises in patients with chronic non-specific low back pain: A review of randomized controlled trials. Niger J Exp Clin Biosci 2019;7:55-62
|How to cite this URL:|
Gujba FK, Lambon N, Ganiyu S, Masta MA, Usman MA. Effectiveness of core stability exercises in patients with chronic non-specific low back pain: A review of randomized controlled trials. Niger J Exp Clin Biosci [serial online] 2019 [cited 2020 Apr 5];7:55-62. Available from: http://www.njecbonline.org/text.asp?2019/7/1/55/266834
| Introduction|| |
Low back pain (LBP) is one of the common major problems that affect adults with about 85% of cases regarded as “nonspecific LBP” due to the nature of the pain being poorly understood., Individual studies have shown that the use of core stability exercises (CSE) alone or combined with other treatment is effective in the management of LBP.,,,,,
The core comprises group of muscles including transversus abdominus, mulfidus, diaphragm, and pelvic floor muscles that supports and stabilises the spine.,,, In LBP, there is limitation of the stabilizing system of the spine to maintain the intervertebral neutral zones resulting in pain, major deformity, and neurological dysfunction, therefore making the spine unstable and cause excessive movements that leads to stretch of muscles, compression of nerves, and inflammation, resulting in LBP.,,
Restricted spinal core stability is one of the predisposing causes of LBP that is recurrent. Instability of the lumbar spine restricts muscle strength, endurance, and flexibility. As a result, attention has been given to CSE programs to improve spinal strength and flexibility, which are mandatory to relieve LBP and continued self-care.,, CSE programs have been carried out widely in the recent years in the management of LBP and have shown to be effective in reducing pain and improving trunk stability.,, However, it is not obviously clear if CSE programs are more effective than the conventional exercise programs in the management of patients with chronic nonspecific LBP (CNLBP).,, Core strengthening with its theoretical basis in the treatment and prevention of LBP helps in decrease of pain and functional improvement in patients with LBP.,,
The aim of this study was to systematically review randomized controlled trials (RCTs) evaluating the effectiveness of CSE in the management of patients with CNLBP.
| Methods|| |
Literature search through databases such as CINAHL, AMED, MEDLINE, PEDro, and Cochrane library was used for the search strategy as they are thought to be the most appropriate for this review. Search terms such as low back pain, core stability exercise, spinal stabilisation, lumbar spine, back exercises, and motor control exercises together with MeSH terms such as TX, TX and MJ in the MEDLINE were used to retrieve more relevant studies on LBP and CSE. The studies were selected according to the inclusion and exclusion criteria to determine the applicability of the review. After the screening, six studies met the inclusion criteria ,,,,, Review of six studies might be enough to provide good evidence for a certain intervention provided the studies are of good quality.,,,, The details of the study selection are provided in [Figure 1].
Studies were included if they were human studies published in English between 2010 and 2015, RCTs, included patients with a diagnosed CNLBP that is confirmed from either the clinical features or radiographic findings (i.e., 12 weeks or 3 month), included patients between the age of 18 and 60 years, focused on CSE, and used either or both visual analog scale (VAS) and Oswestry Disability Index (ODI) as outcome measures.
During screening of the studies extracted from various databases and manual search, studies which did not meet the inclusion criteria of this review above were excluded.
CSE is the exercise that involves the spine and core muscles (mostly the transversus abdominis or multifidus), where the core muscles are tightened to while the spine is being stabilized and then progressed to functional activity.,
Data synthesis involved the combination and summary of findings of the studies selected for the review. The synthesis of the data was done by the descriptive synthesis using the extraction form designed by the reviewer to outline characteristics of the studies.
The selected studies for the review were appraised using the PEDro scale [Table 1]. The methodological quality of all selected studies for the review was strictly assessed by two independent reviewers with blinding.,,,,,
In clinical trials, randomization is very important in assigning participants into groups and helps in providing rigor in the results and data interpretation.,,,, Shamsi et al. assigned the participants into the CSE group and traditional trunk exercise group, Cho et al. randomly assigned participants into CSE group and control group, Zhang et al. and You et al. randomized the participants into two groups experimental and control groups, Hosseinifar et al. assigned the participants into CSE group and McKenzie exercise group, and Muthukrishnan et al. randomized participants into CSE group and conventional physiotherapy group.
Allocation concealment is a way of blinding the group in which participants fall into assisting in minimizing selection and confounding bias in a study.,,, Three studies in this review used concealed allocation process,,, while the other three studies had no allocation concealment.,,
In the presentation of clinical trials, a prognostic variable should be described for each treatment group., Five of the studies included in this review shows patient characteristics of key outcome measures at baseline.,,,, Similarly, in all the included studies of this review within the groups, experimental and control groups baseline characteristics were reported. However, the participants in the study by Shamsi et al. were not similar at baseline. This shows that there was a balance between the statistical groups which strengthened the internal validity of the review.
Blinding in research (participants, therapist, and assessors) is particularly important in the elimination of bias challenges that might occur after randomization.,,,,, In this particular review, four of the included studies blinded their assessors ,,, while the other two studies , did not blind their assessors. Three of the studies ,, did not blind their therapists while the other three studies ,, reported therapist blinding, four of the studies ,,, reported participants blinding while , did not report whether the participants were blinded or not.
Intention-to-treat analysis is done with the intent to eliminate bias which predicts assumed differences among treatment groups.,, Three of the studies performed intention-to-treat analysis ,, while the other three studies ,, did not report whether intention-to-treat analysis was done.
All the included studies reported measure of outcome from 85% of the participants initially allocated to groups. All the six included studies reported the results of between group statistical comparisons (P value) within all the primary outcome measures.
A point measure is considered as a measure of the size of treatment effect. Four of the included studies provides satisfactory level of point measures and measures of variability of their study standard deviations, confident intervals, standard errors, and quartile range,,,, while two of the studies , did not report such.
Eligibility criteria help in providing heterogeneity in a study unless there is potential to enroll the population with certain characteristics. Five of the included study specified their eligibility criteria ,,, while the study by Cho et al. did not specify their eligibility criteria.
| Results|| |
The literature search led to the identification of 136 studies, from which 60 trials were considered as potentially relevant and were retrieved for detailed analysis. After complete reading of 60 trials, 54 were excluded, and finally, six trials met the eligibility criteria ,,,,, [Figure 1].
Design of the included studies
Six trials were used for this review in which two , were RCTs, two , were pragmatic control trials, one was a quasi-RCT, and one  was a block RCT [Table 2].
Participants of the included studies
The entire studies included in the review recruited 479 participants and the articles were published between 2010 and 2015 [Table 2].
Age of participants
Participants in the entire review were within the age range of 18–60 years. Participants were adults with CNLBP in line with the inclusion criteria of this review.
Some studies ,, used both ODI and VAS as outcome measures while others ,, used VAS and other outcome measures.
Core stability exercise intervention
In the study by Shamsi et al., there was a warm-up period with eight stretching exercises and stationary bicycling for 5 min. The intervention in the study by Zhang et al. involved the use of Chinese massage and CSE. In a study, participants performed CSE for 30 min, 3 times a week for 4 weeks. The type of CSE performed was explained. In the study by You et al., the participants in the experimental group were asked to lie in a hook-lying position and a pressure biofeedback unit set at 40 mmHg was placed under the participant's fifth lumbar vertebra to provide visual feedback during the intervention. The study by Hosseinifar et al. included a warm-up period in which the participants exercised on a stationary bike for 5 min and then performed stretching exercises for 10 min before specific exercise were performed. Stabilization exercises were performed in six different levels from easy to difficult. In the study by Muthukrishnan et al., the participants were educated on the anatomy of the core muscles and their function before the start of the program. During the 8-week intervention for the CSE group, there was an emphasis on the contraction of the core muscles lasting for 45 min. The treatment protocol was divided into three phases.
Duration of intervention
The duration of the interventions in the studies included in this review varied from 30 min to 90 min per each session for at least once in a week.
Of the six studies included, three reported dropout of participants during the study period.,, The other three studies reported non withdrawal ,, [Table 3].
| Discussion|| |
This review used a systematic review of RCTs to determine the effectiveness of CSE in the management of CNLBP [Table 3] and [Table 4].
Study  reported that both CSE and traditional trunk exercise reduces pain and disability among CNLBP (P > 0.05) with mean and standard deviation of the VAS score of 15.9 ± 12.4 in the experimental group and control group 14.9 ± 14.1, and ODI was not used in this study. In this study, the result did not show a statistical significant difference among the two exercise groups. The intervention in this study lasted for about 1½ months. However, the duration of the exercise was carried out for 6 weeks with five participants dropping out of the study reason for the withdrawal not stated. This study used a large number of participants and there were blinded and randomized into two groups which has reduced the selection bias that might affect the internal validity of the study. However, the participant's characteristics were not similar at baseline and this could affect the outcome of the study, and despite the dropouts and long duration of the study, measures of outcome were obtained from 85% of the participants which has added to the internal validity of the study.
Study  showed that CSE is effective in treatment of CNLBP when combined with Chinese massage than when Chinese massage alone is used with a strong statistical P value (P < 0.05). VAS score 1.46 ± 0.76 in the experimental group and 2.85 ± 1.58 in the control group, experimental group 13.20 ± 2.42 and control 18.39 ± 3.67 for ODI score. This study had the largest number of participants which adds to part of the strength of the study, and despite the long duration of the intervention, there were only five dropouts in this study. Furthermore, one good measure that has added credibility to the study is the 1-year follow-up that can determine the long-term effect of the exercise. Unlike the study, this study has added significant contribution to the evidence of using CSE for LBP. However, one of the limitations of this study is the fact that the therapist where not blinded and this can impact on the result of the trial, but the participants and the assessors were blinded. Self-reported measures of function were used to measure the physical activity rather than using an objective measure and this might have effect on the final result.
Study  reported that CSE is more effective in the management of CNLBP than routine care (P < 0.05) and VAS score of 24.4 ± 8.7 in the CSE group and 1.5 ± 6.7 in the control group, but ODI was not used in this study. This study had the least PEDro score of 5/11 among the selected studies; however, it has met the eligibility criteria of this review. Participants were similar at baseline and assigned to group randomly, the allocation was not concealed, and this could lead to selection and confounding bias. Unlike the other studies,,,,, there was no blinding of the assessors, therapists, and participants and this may alleviate bias challenges that may occur after randomization. However, there is a good credibility in the sense that there were no dropouts in the study all through the 4 weeks of intervention. However, the study lacked follow-up, therefore; long-term effect of the exercise perceived remained unknown. Eligibility for this study was not specified and this might reduce the external validity of the study.
Study  reported that CSE technique with ankle dorsiflexion is more effective than CSE technique alone in the management of CNLBP with a statistical value (P = 0.001) and VAS score in CSE group with ankle dorsiflexion as 4.25 ± 0.97 and that of the CSE group alone 4.95 ± 0.21 and the mean of ODI in the CSE group with ankle dorsiflexion is 17.90 ± 5.28 and in the CSE group alone is 21.15 ± 1.11. Ultrasound guided visual feedback for assessing the outcomes of the study and this might affect the outcome of the study. In terms of internal validity, unlike the other five studies,,,,, this study has scored 11/11 of the PEDro scale of assessing a study and this has made it a more reliable study. A great limitation that might affect the results of this study is the small sample size for this reason the result cannot be generalized to all LBP patients. This study had a long duration of intervention of about 8 weeks, but the fact that there were no dropouts had added quality to the study. The 2-month follow-up in the study has made the study more reliable and has pointed out some evidence of using CSE for LBP.
Study  reported that CSE exercises are more effective than McKenzie exercises in CNLBP management with a significant statistical P < 0.05 and a VAS score of 1.53 ± 1.40 in the CSE group and 2.66 ± 1.39 in the McKenzie group, ODI being not used. Participants were randomly allocated to groups, and as similar to the study, above allocation was not concealed which might cause selection bias in the trial. The intervention was carried out for 6 weeks. There were three withdrawals due to participation decline this could have impacted on the external validity of the study, but the fact that measures of outcome were obtained from 85% of the participants has made the validity of the study less threatened. Despite the dropouts, there was no intention-to-treat analysis. The result of this study can be accepted because the eligibility criteria has been specified and the participant's characteristics were similar at baseline. Also in this study the point measures of variability were clearly reported which made the result of the study much stronger.
Study  reported that CSE regimen is more effective than the conventional physiotherapy regime in the management of CNLBP (P < 0.05) and VAS score in the conventional physiotherapy regimen group is 3.2 ± 1.2 and that of the CSE group is 3.9 ± 0.8, ODI being not used. To ensure internal and external validity of this study, they randomized their participants into groups with concealed allocation to minimize bias in the study. The result of this study also gained credibility due to proper blinding of the assessors, therapists, and participants and also because there were no drop outs in a longitudinal study that had a long duration of about 8 weeks. They intentionally avoided the use of large sample size in this study to avoid large clinical load and this can add to the quality of the study. Despite the small sample size in the study, they failed to report the point measure of variability in the study which if reported might add to the internal validity of the study.
All the six studies critically appraised above have sound internal validity which makes the results of the studies acceptable and can be used as evidence in future practice.
A limitation of this review is inability to include other studies that were reported in other languages other than English and age range of the participants was restricted to 18–60 years only. Another limitation is that PEDro scale was the only scale used for the assessment of the internal validity of the studies. Furthermore, we included only studies that used VAS and ODI as outcome measures.
| Conclusion|| |
The results of this review support the effectiveness of CSE for the management of pain and disability among CNLBP patients. This was shown in five out of the six studies that were systematically reviewed. However, evidence to support the long = term effect of the CSE intervention on CNLBP is weak. This is because statistically significant improvement after follow-up was only reported in two of the studies included in this review. Although our review might not have revealed all about CSE and CNLBP, the findings might assist health-care professionals understand and appreciate the use of CSE in the management of CNLBP.
Implications for future research
The result of this study could be disseminated to other health-care professionals through publications in journals. A longer duration of time and more number of studies should be used. Future reviews in other to ascertain the effectiveness of the findings need to be conducted.
Future studies should adopt robust methodology to assess the long-term effects of CSE in the management of CNLBP. In addition, future studies should look at the effects of individual CSE for other LBP other than CNLBP. Other assessment scales should also be used for assessing the internal validity in the future study. Furthermore, non-English studies should also be included if it is understandable for the reviewer.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest
| References|| |
Ahmed R, Shakil-Ur-Rehman S, Sibtain F. Comparison between specific lumber mobilization and core-stability exercises with core-stability exercises alone in mechanical low back pain. Pak J Med Sci 2014;30:157-60.
Searle A, Spink M, Ho A, Chuter V. Exercise interventions for the treatment of chronic low back pain: A systematic review and meta-analysis of randomised controlled trials. Clin Rehabil 2015;29:1155-67.
Hwangbo G, Lee CW, Kim SG, Kim HS. The effects of trunk stability exercise and a combined exercise program on pain, flexibility, and static balance in chronic low back pain patients. J Phys Ther Sci 2015;27:1153-5.
Cho HY, Kim EH, Kim J. Effects of the CORE exercise program on pain and active range of motion in patients with chronic low back pain. J Phys Ther Sci 2014;26:1237-40.
Sokunbi O, Muhwhati L, Robinson P. A pilot study on using acupuncture and core stability exercises to treat non-specific acute low back pain among industrial workers. S Afr J Physiother 2014;70:4-10.
Zhang Y, Tang S, Chen G, Liu Y. Chinese massage combined with core stability exercises for nonspecific low back pain: A randomized controlled trial. Complement Ther Med 2015;23:1-6.
Wang XQ, Zheng JJ, Yu ZW, Bi X, Lou SJ, Liu J, et al.
A meta-analysis of core stability exercise versus general exercise for chronic low back pain. PLoS One 2012;7:e52082.
Choi BK, Verbeek JH, Tam WW, Jiang JY. Exercises for prevention of recurrences of low-back pain. Occup Environ Med 2010;67:795-6.
Akuthota V, Nadler SF. Core strengthening. Arch Phys Med Rehabil 2004;85:S86-92.
Lederman E. The myth of core stability. J Bodyw Mov Ther 2010;14:84-98.
McGill SM. Low Back Disorders. 3rd ed. Vol. 13. UK: Human Kinetics; 2015. p. 9781450472913.
Dunsford A, Kumar S, Clarke S. Integrating evidence into practice: Use of McKenzie-based treatment for mechanical low back pain. J Multidiscip Healthc 2011;4:393-402.
Cox JM. Low back pain: mechanism, diagnosis and treatment. Vol. 13. US: Lippincott Williams and Wilkins; 2012. p. 978-1608310029.
Akuthota V, Ferreiro A, Moore T, Fredericson M. Core stability exercise principles. Curr Sports Med Rep 2008;7:39-44.
Schembri L, Fenech P, Sacco, M. Low back pain: A Comparative study on the value of core training vs traditional strengthening exercises. Mjhs 2014;1:12-18.
Ganiyu SO, Gujba KF. Effects of acupuncture, core-stability exercises, and treadmill walking exercises in treating a patient with postsurgical lumbar disc herniation: A clinical case report. J Acupunct Meridian Stud 2015;8:48-52.
Sokunbi O, Maduagwu S, Kachalla FG. The use of alternate side lying manipulation, acupuncture and core stability exercises in the treatment of multiple level disk prolapse a case study. J Dent Med Sci 2014;13:90-6.
Higgins JP, Green S. Cochrane Handbook for Systematic Reviews of Interventions. West Sussex: John Wiley & Sons; 2008.
Rothwell PM. External validity of randomised controlled trials: “To whom do the results of this trial apply?”. Lancet 2005;365:82-93.
Shamsi MB, Sarrafzadeh J, Jamshidi A. Comparing core stability and traditional trunk exercise on chronic low back pain patients using three functional lumbopelvic stability tests. Physiother Theory Pract 2015;31:89-98.
You JH, Kim SY, Oh DW, Chon SC. The effect of a novel core stabilization technique on managing patients with chronic low back pain: A randomized, controlled, experimenter-blinded study. Clin Rehabil 2014;28:460-9.
Hosseinifar M, Akbari M, Behtash H, Amiri M, Sarrafzadeh J. The effects of stabilization and McKenzie exercises on transverse abdominis and multifidus muscle thickness, pain, and disability: A randomized controlled trial in nonspecific chronic low back pain. J Phys Ther Sci 2013;25:1541-5.
Muthukrishnan R, Shenoy SD, Jaspal SS, Nellikunja S, Fernandes S. The differential effects of core stabilization exercise regime and conventional physiotherapy regime on postural control parameters during perturbation in patients with movement and control impairment chronic low back pain. Sports Med Arthrosc Rehabil Ther Technol 2010;2:13.
Duthey B. Priority Medicines for Europe and the World: A Public Health Approach to Innovation. WHO Background Paper 6; 2013.
Kabisch M, Ruckes C, Seibert-Grafe M, Blettner M. Randomized controlled trials: Part 17 of a series on evaluation of scientific publications. Dtsch Arztebl Int 2011;108:663-8.
Tooth L, Bennett S, McCluskey A, Hoffmann T, McKenna K, Lovarini M. Appraising the quality of randomized controlled trials: Inter-rater reliability for the OTseeker evidence database. J Eval Clin Pract 2005;11:547-55.
Smith BE, Littlewood C, May S. An update of stabilisation exercises for low back pain: A systematic review with meta-analysis. BMC Musculoskelet Disord 2014;15:416.
Coulombe BJ, Games KE, Neil ER, Eberman LE. Core stability exercise versus general exercise for chronic low back pain. J Athl Train 2017;52:71-2.
Prictor M, Hill S. Cochrane consumers and communication review group: Leading the field on health communication evidence. J Evid Based Med 2013;6:216-20.
Jensen MP, Karoly P, O'Riordan EF, Bland F
Jr., Burns RS. The subjective experience of acute pain. An assessment of the utility of 10 indices. Clin J Pain 1989;5:153-9.
Davidson M, Keating JL. A comparison of five low back disability questionnaires: Reliability and responsiveness. Phys Ther 2002;82:8-24.
Verhagen AP, de Vet HC, de Bie RA, Kessels AG, Boers M, Bouter LM, et al.
The Delphi list: A criteria list for quality assessment of randomized clinical trials for conducting systematic reviews developed by Delphi consensus. J Clin Epidemiol 1998;51:1235-41.
Maher CG, Sherrington C, Herbert RD, Moseley AM, Elkins M. Reliability of the PEDro scale for rating quality of randomized controlled trials. Phys Ther 2003;83:713-21.
Maxwell L, Santesso N, Tugwell PS, Wells GA, Judd M, Buchbinder R. Method guidelines for Cochrane musculoskeletal group systematic reviews. J Rheumatol 2006;33:2304-11.
Scott NW, McPherson GC, Ramsay CR, Campbell MK. The method of minimization for allocation to clinical trials. A review. Control Clin Trials 2002;23:662-74.
Kalish LA, Begg CB. Treatment allocation methods in clinical trials: A review. Stat Med 1985;4:129-44.
Schulz KF, Grimes DA. Allocation concealment in randomised trials: Defending against deciphering. Lancet 2002;359:614-8.
Suresh K. An overview of randomization techniques: An unbiased assessment of outcome in clinical research. J Hum Reprod Sci 2011;4:8-11.
] [Full text]
Schulz KF. Randomised trials, human nature, and reporting guidelines. Lancet 1996;348:596-8.
Dykiert D. Data extraction, quality assessment and narrative synthesis. Centre for cognitive ageing and cognitive epidemiology. J Am Med Assoc 2000;283:2008-12.
Altman DG. Better reporting of randomised controlled trials: The CONSORT statement. BMJ 1996;313:570-1.
Senn SJ. Covariate imbalance and random allocation in clinical trials. Stat Med 1989;8:467-75.
Karanicolas PJ, Bhandari M, Taromi B, Akl EA, Bassler D, Alonso-Coello P, et al.
Blinding of outcomes in trials of orthopaedic trauma: An opportunity to enhance the validity of clinical trials. J Bone Joint Surg Am 2008;90:1026-33.
Devereaux PJ, Bhandari M, Clarke M, Montori VM, Cook DJ, Yusuf S. Need for expertise based randomised controlled trials. BMJ 2005;330:88.
Sheiner LB. Is intent-to-treat analysis always (ever) enough? Br J Clin Pharmacol 2002;54:203-11.
Moncur R, Larmer J. Clinical applicability of intention-to-treat analyses. Evid Bas Med 2009;6:39-41.
Gupta SK. Intention-to-treat concept: A review. Perspect Clin Res 2011;2:109-12.
] [Full text]
Bell DL, Breland DJ, Ott MA. Adolescent and young adult male health: A review. Pediatrics 2013;132:535-46.
Wood AM, White IR, Thompson SG. Are missing outcome data adequately handled? A review of published randomized controlled trials in major medical journals. Clin Trials 2004;1:368-76.
[Table 1], [Table 2], [Table 3], [Table 4]