The title of this piece is concise, to the point and easy to understand. It is entirely appropriate for the paper. The only difficulty is that it implies a direct comparison between plain X-Ray films and MRI scans which, it later becomes apparent, it is not.
There are no key words offered in this paper
The abstract in this paper is caught between being concise and giving sufficient information to the reader. The issues raised in this study are complex and not easily condensed into a couple of sentences. The authors in this case err on the side of offering a long abstract but do cover (nearly) all of the main points. They miss out the arguments put forward at some length in the actual study relating to the importance of potential litigation in the decisions about imaging. They do, however, set out the main points of the study together with the study design and, to a limited extent, also set out the prime findings of the study. This appears to be entirely appropriate for this type of paper.
The introduction of this piece is comparatively short as the literature review essentially forms part of the introduction but it is set out as a separate segment.
The introduction segment offers a brief overview of the incidence and clinical significance of scaphoid fractures. The facts are appropriately supported with good quality references from the literature. The points made which contrast the modalities of X-Rays and MRI scans in terms of quality, expense and detection rate are succinctly made and quite appropriate.
It offers a brief rationale of the project by setting out the difficulties with plain X-Ray films of suspected scaphoid fractures and then sets out the advantages and disadvantages of MRI scans for the same condition. The introduction closes with what is essentially the core of the paper, the definition of the “gold standard” for medical investigation and follows this with the two questions which explicitly form the basis of the whole study. (Gomm R et al. 2000). On reading this segment of the introduction, one could be forgiven for concluding that the two definitions of the “gold standard” are cited from another authority as the authors use the term “is regarded” which suggests that it is regarded by others. As no citation is given, one should therefore properly conclude that the definition is one suggested by the authors. (Polit D F et al. 2004). This is quite an important point as the respondents are being asked to consider if plain X-Ray films are to be considered “the gold standard” and it is entirely possible that they may choose to define it with different criteria.
The literature review
This segment effectively forms the basis of the introduction and is extensive. It is well constructed, logically set out and carefully explained. The authorities utilised are however, rather disappointing as they are primarily old papers. (Bowling A 2002).
It is accepted that this article was written in 2005, even though it was not published until 2007, but the vast majority of the authorities cited are from the last decade with some dating from 1991. If one looks at even a brief literature search on the subject of X-Rays, MRI scans and scaphoid fractures, it will be quickly seen that there are a great many good quality authoritative papers which could perhaps have been more appropriately cited and are far more recent (viz. Tan S et al. 2009).The segments on the clinical difficulties and rational management are particularly well set out offering a balanced argument. The same can be said of the segment on the use of plain X-Ray films where the clinical benefit is traded off against the extra exposure to ionising radiation.
As the thrust of the investigation is whether plain X-Ray films or MRI scans are better in the long term, it would have been useful to know the degree to which the ionising radiation used in multiple (usually four) wrist X-Rays is likely to cause harm, if that is known, as the study later weighs up the detrimental effect of plain X-Ray films against the positive features of the MRI scans which are said not to cause damage.
Another element which could have been improved is the segment which deals, very commendably, with the different types of scaphoid fracture. The relevance of this is that the authors state that different types of scaphoid fracture have different degrees of difficulty of interpretation on plain X-Ray films. This literature review does not give any information as to whether it is the commonest form of scaphoid fractures that is hardest to spot or the rarest. Such information clearly has relevance when considering the value of the MRI scan.
A puzzle was the inclusion of tomographic examinations which, the authors suggest, are better than plain X-Ray films for diagnosing scaphoid fractures. One wonders then why these were not included in a comparative study since, intuitively, they would have similar disadvantages to plain X-Ray films but if they have a higher detection rate, they may be considered better than the more expensive and restrictive MRI scans. (Denscombe M 2002)
In essence, the literature review clearly depicts a gap in the literature which this study seeks to address.
MATERIALS & METHODS
The methodology in this study was actually rather simple. It involved 100 questionnaires sent to 20 lead radiologists at 20 local hospitals with a request to “cascade” four other questionnaires to their colleagues. A technique known as ‘convenience sampling'. (Fowler, F. J. 2000). One slightly troubling element was that the authors state that “an attempt was made to determine if MRI facilities were available.” This is an odd admission, since if MRI scans were not available to any of the radiologists, then this would be a clear and huge source of respondent bias. (Schulz K F et al. 1995). One might be forgiven for thinking that instead of “making an attempt” to find out, it might have been better if the authors had actually found out this fairly crucial piece of information.
The choice of semi-structured questionnaire is quite appropriate for this type of investigation. It offers comparative ease in assembling large quantities of relevant information as well as allowing the respondents to add additional information or amplification if they felt it was appropriate. This additional information was analysed using standard pre-constructed response grids. (Patton, M Q. 2000)
No information is given relating to the decisions of optimum sample size or power calculations, but then one would have to concede that the statistical analysis was rudimentary at best and statistical significance was not attempted. (Rosenthal R. 1994)
This is undoubtedly the weakest segment of the paper. The results, particularly the graphical results, are confusing and not easy to understand without considerable reader effort. In Fig 1. The use of two separate bars to illustrate the same distribution seems completely unnecessary and confusing. Table three presents a lot of information in a manner which is difficult to extrapolate from the diagram. One particularly irritating feature is the fact that the authors switch from citing results as numbers (n) to percentages and back rather than offering a consistent presentation throughout.
The fact that the questions are actually set out is commendable as frequently papers are read where one is left to assume exactly what questions have been shown to the respondents and it is only by knowing what was asked that one can accurately judge if the questions have been answered appropriately.
The fact that one centre did not use MRI scans for scaphoid fracture investigation puts an unacceptable element of bias in the results. How can the researchers expect consultants who don't have the option of using a MRI scan to offer an opinion which can be included with those of consultants who routinely do use them? (Schulz K F et al. 1995).
After reading the introduction, literature review and methodology segments, one would have expected an altogether more robust results section.
The discussion element of the paper essentially highlights the difference between the theoretical and practical elements of the situation. The authors clearly are writing this element of the paper from the viewpoint that the MRI scans should be considered the “gold standard” for the diagnosis of scaphoid fractures. The first comment in the segment “MRI” shows that they consider that the case has already been made by the literature review. While that may indeed be the case, many of the respondents simply did not use the MRI scans for a number of reasons. (Hansen T B et al. 2009)
There is no merit in simply rehearsing all of the arguments put forward in this section. One can critically observe that it is the probably the most effectively written element of the paper and it covers all elements of the study, not only the results of the questionnaire alone.
A feature which arose from the results and was commendably included in the discussions was that a significant number of the respondents actually thought that neither plain X-Ray films or MRI scans should be considered the gold standard but a number of (unconsidered) modalities viz, CT, ultrasound and RNI.
These options were not explicitly put to the other respondents and therefore may well have diluted the opinions that were expressed if they had been. (Mohammed, D et al. 2003)
The authors appear to have offered a frank and considered overview of the whole project in their discussion segment with no obvious exaggeration or overt bias.
It is also commendable that the authors refer back to the literature to underpin and to illustrate some of the points made in the discussion.
As has already been highlighted, the fact that one centre did not use MRI scans for scaphoid fracture investigation, for reasons that were not made clear, puts a completely unacceptable element of bias in the results. (Wood M 2003). It is not appropriate for the researchers to expect consultants who don't have the clinical option of using a MRI scan to offer an opinion which is then statistically compared with the opinions of consultants who routinely do use them.
The original aims of the study, that it should determine whether professionals considered MRI scans to be the gold standard for the investigation of scaphoid fractures was achieved, to a degree, within the limitations set out above.
The conclusions in this study are set out concisely and appear to be appropriately deduced from the results obtained from both the literature review and the questionnaire elements of the study. It does stray from the research question when it starts to discuss the issues of whether one is over treating a lot of patients in order to avoid under treating a few, but although not explicitly set out in the rationale, it is still an important point which arises from the data derived from the study
The actual research question “Do professional still perceive plain X-Ray films to be the gold standard in the detection of scaphoid fractures?” is not actually answered, as the conclusion segment actually is a setting out of the supporting and refuting elements of the answer. One would have to deduce that the answer to the question is “Some still do”. There is no recommendation for further research in this area.
The reference segment is actually quite extensive for a study of this nature. Almost without exception they are primary sources. The downside is that many of them are really rather old. (Bell J 1999). Even a brief literature search of the peer reviewed journals on this subject shows a wealth of articles and studies which could have been more usefully used as they are considerably more recent. (Robson C 2002)
An overview of the study shows a commendable effort which clearly started off with a clear direction. The work up and the literature review is excellent but the presentation of the results let it down. This is disappointing considering the obvious effort that has been expended on many of the design and execution elements of the study.
The reader is however, able to come away with a clear and unequivocal impression of the main arguments and also to draw their own conclusions from a critical analysis of the study. (Carr LT 1994)
It adds appreciably to the evidence base in this area.
Bell J (1999) Doing your research project. A guide for first time researchers in education and social science. 3rd edition. Buckingham; Open University Press.
Bowling A (2002) Research Methods in Health: Investigating health and health services. (2nd Edn) Open University Press, Maidenhead.
Carr LT (1994) The strengths and weaknesses of quantitative and qualitative research: what method for nursing? Journal of Advanced Nursing Volume 20 Page 716 - October 1994
Denscombe M (2002) Ground rules for good research. Maidenhead, Open University Press.
Fowler, F. J. (2000) Survey research methods. 3rd ed. London: Sage.
Gomm R & Davies C (2000) Using evidence in health and social care. London, Sage / The Open University.
HANSEN T B, PETERSEN R B, BARCKMAN J, UHRE P, ARSEN K (2009) Cost-effectiveness of MRI in managing suspected scaphoid fractures J Hand Surg Eur Vol., October 2009; 34: 627 - 630.
Mohammed, D Braunholtz, and T P Hofer (2003) The measurement of active errors: methodological issues. Qual. Saf. Health Care, Dec 2003; 12: 8 - 12.
Patton, M.Q. (2000) Qualitative Evaluation and Research Methods. Thousand Oaks, California, USA: Sage Publication
Polit DF & Beck CT (2004) Nursing Research: Methods, appraisal and utilisation (6th ed.) Philadelphia; Lipincott.
Robson C. Real World Research. Oxford: Blackwell; 2002.
Rosenthal R. (1994) Parametric measures of effect size. In: Cooper H, Hedges LV, eds. The handbook of research synthesis. New York: Russell Sage Foundation, 1994.
Schulz K F, Chalmers I, Hayes R J, Altman D G. (1995) Empirical evidence of bias. Dimensions of methodological quality associated with estimates of treatment effects in controlled trials. JAMA 1995; 273 : 408 - 412
Tan S, Craigen M A C, Porter K (2009) Acute scaphoid fracture: a review. Trauma, October 2009; 11: 221 - 239.
Wood M (2003) Making sense of statistics: a non-mathematical approach. Oxford: Blackwell; 2003.