The Validity of Mandatory Follow-Up Skeletal Surveys - Essay Example

The Validity of Mandatory Follow-Up Skeletal Surveys Contents Page Page Glossary 1.0 Introduction 3-7 2.0 Literature Review 7-15 3.0 Method 15-18 5.0 Discussion 18-22 References 23-26 Appendix A 27 Appendix B Child abuse casts a shadow the length of a lifetime. - Herbert Ward (1916; British artist) Assessing the Validity of Mandatory Follow-Up Skeletal Surveys 1.0 Introduction Aim Identification of factors that influence the use, and non-use, of skeletal survey as a mandatory tool for suspected child abuse cases in the UK. Non-accidental Injury (NAI) in children ( for the purpose of this study a child is 2 years of age or less) indicates severe assault, and must be identified if present (Kemp, Butler, Morris, Mann, Kemp, Rolfe, Sibert, & Maguire, 2006). Radiology plays a crucial role in the diagnosis of the sensitive subject of suspected NAI (Offiah, & Hall, 2003) as records show that 55% of physically abused young children (3 years old) have fractures with 80% of this group being less than 18 months of age (Kemp et.al, 2006). Most often, according to Thompson (2005), physical child cruelty has little or no visible evidence, with pre-verbal children at most risk of being victims with injuries that are frequently at multiple locations, with varying degrees of a healing stage and in several organ systems. For the duration of infancy, mortality is greater through injury than all other combined causes opine (Mckinney, Lane, & Hickey, 2004). Dr John Caffey's (ref) [this is your citation, not mine …?] identification of unexplained fractures (mid 1940s) in children who presented to accident and emergency (A+E) departments has led to radiographic technological advances and led to the conceptualisation of “the battered child syndrome”. Injuries to the skeleton are important markers in physical child abuse. Fractures with a high specificity for maltreatment have been described in many sections of the axial and appendicular skeleton (Kleinman, Marks, Richmond, & Blackbourne, 1995). Identification and dating of skeletal injuries are key factors in alleged child abuse according to Kleinman and colleagues (1995) with the radiographic findings frequently being used in criminal proceedings. As the incidence of child abuse is of a higher incidence that that of genetic or metabolic bone diseases, differential diagnosis of multiple fractures must have child abuse at the forefront of investigation. A follow-up skeletal survey provides a way to carefully gather imaging data to reinvestigate old and new fractures that may indicate inflicted trauma (Block, 2005). Several studies attest to the viability of using skeletal survey and follow-up skeletal survey to assess and confirm suspected cases of NAI (Ellerstein & Norris, 1984; Kemp et al., 2006; Nimkin, Spevak, & Kleinman, 1997; Kleinman, Kleinman, & Savage, 2004). The skeletal survey itself is a systematic series of radiographic images that picture the entire skeleton or at least the anatomic areas that are considered appropriate to indicate a clinical condition, as shown in Table 1 (see Appendix A).The skeletal survey aims to identify with precision the focal and diffuse abnormalities that may exist within the skeleton. These include fractures which may be healing, as well as fractures across a range of ages. The survey aids in the assessment of differentiating developmental changes from other anatomical changes that may occur for a child or infant. There are several indications for a skeletal survey: 1) For suspected physical abuse of infants and children; 2) For suspected skeletal dysplasias, syndromes and metabolic disorders; 3) Suspected neoplasia and related disorders. Several guidelines exist to aid radiologists in the effective and efficient use of radiology for young patients. However, in the UK there is no accepted policy of who and when a skeletal survey should be undertaken, unlike in the USA (Day, 2005). At present there are standards in draft from the British Society for Paediatric Radiology (BSPR) and this aims to define indications for skeletal survey when NAI is suspected. Meanwhile, the application of standards remains diverse throughout the UK. It is advisable though, with the skeletal survey to radiograph at least the frontal projection. Those of the axial skeleton are advisable in at least two projections. Using oblique projections to the anteroposterior (AP) view of the rib cage is advisable to increase the possibility of detecting a fracture in that area. In order to gain a comprehensive record of suspected injuries it is usually necessary to radiograph from several angles, such as; an oblique of the skull; centred at the joints; and lateral views of the extremities. This research study aims to identify factors which indicate the viability of using skeletal survey as an initial assessment tool, and follow-up survey as a method of confirmation of suspected NAI. To achieve this end, this study had several objectives: 1) To determine characteristics which make the skeletal survey an obvious choice for follow-up assessment? 2) To review current discourse and literature as to the use of follow-up skeletal surveys in the assessment of suspected child abuse. It is anticipated that the following dissertation shall answer the primary research question: Should the follow-up skeletal survey be mandatory in all cases of suspected child abuse in infants? 2.0 Review of the Literature Article 1: Kleinman, P. L., Nimbin, K., Speak, M. R., Raider, S. M., Madams. M. L., Shelton, Y. R., Patterson, M. M. (1995) Follow-up skeletal surveys in suspected child abuse. AJR, 167: 893-896. Lineman et al (1995) conducted a retrospective study in the USA to investigate the use of a repeat skeletal survey in the identification and dating of skeletal injury of children where physical abuse was highly likely to have occurred. The purpose of the study aligned with the intentions of the present study. The literature review was scarce, although the studies contained were relevant and up-to-date. Also, the hypothesis was not clearly defined, with the authors stating only that the study was to evaluate the additional yield that a follow-up skeletal survey would provide; hence they did not make a hypothesis to test against their data. A reasonable hypothesis would have been that their study was anticipated to reveal an increase in fractures, or particular fractures as compared to those diagnosed in the initial survey. This would have allowed an ANOVA to be used to compare the groups, rather than the non-parametric Chi-Square significance test. As such, the results would have had higher internal reliability and external validity because the range of initial assessments would have meant diversity in measurement consistencies, across GPs, other imaging techniques, and family and personal medical histories. The pre-post design used a sample consisting of 23 infants and toddlers who were suspected to have been recently physically abused as inferred from initial findings on a skeletal survey and other imaging techniques, as well as family and medical history and or physical examination. The design was appropriate in that it provided a comparison group for the study. However, that a range of initial diagnosis sources were used made it more difficult to compare the results of this study, it would have been more appropriate to have made the different initial assessments different groups, and then compare the results of each group against the follow-up outcomes. The sample was small in that it is unlikely that it was a normal distribution and so unlikely to reflect the population it was to represent and so lacked external validity. The age range did not was not representative of children, as it was limited to infants and toddlers. The sample could have been larger if it included older children (i.e., up to 10 years old) and would have provided a comparison group for the infants and toddlers. No mention was made of informed consent being obtained or how this occurred if it did. A fortnight after the first examination, a follow-up examination was given using the same high-detail imaging The skeletal survey procedure required tightly collimated anteroposterior views of the appendicular skeleton and anteroposterior and lateral views of the axial skeleton. It is clear that the procedure did not meet the complete draft guidelines suggested by the BSPR, as not all the possible radiographic angles were undertaken. The results showed that the second skeletal survey provided additional information about injury in 61% of the participants; as an example shows in Figure 1, (see Appendix A). The second examination also revealed an increased number of definite fractures (27%) (p = .005). Also, the majority of the additional fractures detected (94%) were classic metaphyseal lesions, otherwise known as rib fractures. Article 2: James, S. L., Holliday, K., Somers, J., & Broderick, N. (2003) A Survey of Non-accidental Injury Imaging in England, Scotland and Wales. Clinical Radiology, 58(9): 696-701. A postal survey was used for a case control design across England, Scotland and Wales to investigate national variation in NAI imaging (James, Halliday, Somers & Broderick, 2003). The purpose of this study partially aligned with the outcomes of the present study in that id did explore the extent to which follow-up skeletal surveys were used. The literature review was both comprehensive and up to date. This method has the lowest response rate of all survey methods and it is difficult to confirm who has actually completed the survey. However its advantages were that it was low cost and that it provided wider access to the population as a sample could be gathered across the UK as opposed to a local region, thus providing a better idea of what is happening in different places. This method also aided with the anonymity of respondents as it is unlikely that the researchers had any direct contact with them. Also, as this method is highly convenient for the participants, answers were likely to be more honest, and not rushed. The overall sample was comprised of; 21 teaching hospitals (22%); eight children's hospitals (8%); and 68 district general hospitals (70%). The sample was quite large which would have increased the likelihood of high external validity as the sample could be considered representative of the types of hospitals that would use follow-up skeletal survey to identify NAI. However, as the results were compared across these groups it would have increased the internal reliability if more children's hospitals had been contacted to participate. No mention was made of informed consent being obtained or how this occurred if it did. Overall, participants stated that there was a wide range of variation for all forms and elements of NAI imaging as well as the methods used. There was also a large amount of diversity in the total number of cases, and use of follow-up imaging, including that of skeletal survey. However, this was to be expected given that no national protocol exists that is inclusive of all elements of NAI imaging. Responses showed that district general hospitals had NAI imaging that was the reporting responsibility of general radiologists, whereas reporting at teaching or children's hospitals was the responsibility of a specialist. There was also a range of responses for follow-up imaging. Only 25 of the hospitals routinely used follow-up imaging, and 23 used the procedure occasionally on individual case merit. As shown in Table 2 (see Appendix A), the use of follow-up skeletal survey occurred for only 2 of the 25 hospitals responding that they use any imaging technique at all. Hence, this study was not the most appropriate to answer the research question for the present study. With regard to skeletal survey protocols of the sample, their procedures were compared with the draft BSPR document for suspected NAI. It was found that all of the hospitals in the sample followed the BSPR guidelines with regard to lateral skull radiography and the frontal chest radiograph. Also, 72% followed the guidelines with regard to AP skull radiograph, whereas only 20% followed advice to perform oblique views of the ribs. Only basic descriptive analysis was carried out on the survey data, and so the external validity was very low, as not even non-parametric tests were used to test for significance. The sample was small in that higher order statistics could not be used, for that the sample would need to be around 300 to provide a questionnaire that could be analysed with factor analysis (to determine factor loadings), or regression analysis to see relationships and make predictions. Article 3 Block, R. W. (2005) Follow-up skeletal surveys prove to be valuable in evaluation of child physical abuse. Child Abuse Neglect, 29(10): 1073-1074. This study used a retrospective study of a pre-post design for a sample of 26 UK children who had been recommended for follow-up skeletal imaging. The purpose of the study complemented the present study. The study did not clearly state its hypothesis and the literature review was quite sparse and outdated considering the number of studies that had previously been undertaken on this topic. The method was inadequate in that a follow-up skeletal survey was undertaken for children under the age of 36 months but again a range of radiographic techniques and or general practitioner examination were the comparison groups. As stated previously, this did not provide an ideal comparisons group due to the variability of standards and people carrying out the initial assessments. This would have seriously affected the internal and external reliability of the study, as replication of the study could not assure that similar results would be found if this method was followed. No mention was made of informed consent being obtained or how this occurred if it did. It was also limitation for this study to have such a small sample, as this has lowered the power of the study. So that although it was found that having a follow-up survey was of great use to confirming diagnosis of NAI, it may be that the small sample size skewed the results due to it being unlikely that 26 participants reflect a normal distribution. As such, any difference between the initial and follow-up survey could be magnified due to the use of a non-parametric test that is Chi Square test of significance. As such, the test provides a rough estimate of confidence in determining if two sets of data are different enough in their characteristics, so that the results can be generalised from the sample to a lager population. As a non-parametric test the chi-square test of significance is a rough estimate of confidence in the results. This is because the test has more flexible assumptions, such as being able to use data that is not normally distributed. However, there is a need for the data to have at least 5 cases in each cell to run the analysis; a sample of 26 would provide this. However, the follow-up survey was found to be of benefit in providing more detail for the cases included, and aided in the tentative diagnosis of abuse to be confirmed, or excluded as the cause of trauma. The researchers suggested that a larger study be used in future to determine the need for follow-up skeletal surveys in only children and infants who present with fractures of tentative fractures (i.e., questionable bone findings). Article 4 Day, F., Clegg, S., McPhillips, M., & Mok, J. (2006) A retrospective case series of skeletal surveys in children with suspected non-accidental injury. Journal of Clinical Forensic Medicine, 13(2): 55-59. A more recent retrospective study by Day and colleagues in 2005 investigated the diagnostic efficiency of skeletal surveys in general, and identifying their predictive validity with regard to clinical practice. This pre-post design study reflected the purpose of the present study in that it investigated follow-up skeletal surveys. And their literature review was comprehensive, concise and up to date, clearly identifying gaps in knowledge to justify their design. Their study did not require ethical approval, though the authors were required to complete a data protection form, as per local policy. The setting was a children’s hospital in the very north of England, and the studies were for suspected NAI. It was a benefit to this study to detail their ethical considerations as it aids future research in this area. The surveys were conducted at the Royal Hospital for Sick Children in Edinburgh, and were for suspected NAI. All cases between 01/01/99 and 31/12/03 were reviewed. The purposive sample consisted of 77 children, 6 of these being siblings. Of the cases, 24% of the skeletal surveys were positive, having a mean 2.5 fractures per child. The ages of positive cases ranged from 2 weeks to 36 months, the majority (82%) being under 12 months of age. The indications for skeletal surveys were head injury (12%), intracranial injury (24%), skeletal fracture (35%) and bruising (24%). However a Chi square significance tests showed no significant differences in age, sex or primary reason for skeletal survey between those with a negative and positive skeletal survey. The occult injuries are listed in Table 3 (see Appendix A). The tibial fractures comprised of metaphyses in 4 of the cases. One of the children had a bilateral femoral shaft fracture. And although 53% of the children had only one fracture, 18% had more than 5. Of the three, 77% of their injuries were rib fractures. The authors concluded that the use of skeletal surveys was critical in the ascertaining of incidence of NAI. However the sample was small, and there was no follow-up skeletal survey conducted, or complementary radiography imaging used. Small sample numbers could have provided the Chi-Square test of significance with the ability to detect a difference. An alternative and much more stringent method of collecting reliable and valid data would have been to randomly assign participants to three groups; a complementary follow-up radiograph (e.g., skeletal scintigraphy); follow-up skeletal survey; and physical assessment by a GP alone. This would have enhanced the generalisability of the study as it would allow for parametric statistical testing, such as an ANOVA (Analysis of Variance). Article 5 Zimmerman, S., Markoroff, K., Care, M., Thomas, A., & Shapiro, R. (2005) Utility of follow-up skeletal surveys in suspected child physical abuse evaluations. Child Abuse & Neglect, 29(10): 1075-1083. Also in 2005 was the prospective pre-post study conducted by Zimmerman and colleagues to investigate the viability of using follow-up skeletal surveys in cases of suspected child abuse (Zimmerman, Markoroff, Care, Thomas, & Shapiro, 2005). As such, the purpose of this study aligned with the needs of the present study. The literature review was sparse in that it did not clearly support the design chosen for the study, however the references used were up to date. Follow-up surveys were recommended for 74 infants and toddlers following an initial assessment by a Child Abuse Team in Cincinnati, USA. It was a limitation of this study that the age range, and gender break-up of the sample were not reported to aid in furthering the understanding of this population. No mention was made of informed consent being obtained or how this occurred if it did. The surveys took place at Children's Hospital Medical Center where only one paediatric radiologist (MC) who was a member of the Child Abuse Team read all of the initial and follow-up skeletal surveys in this study. This increased the internal and external validity of the study in that demand characteristics due to the gender, age or ethnicity of the researcher were limited to just one person. The MC was also given access to the history, and records of participant's physical examination findings, and other assessments by the Child Abuse Team. This increased the internal reliability of the study in that there would have been more consistency across the MC judgements of the different records and skeletal survey radiographs. There would also have been reduced risk of testing threat (i.e., the act of the initial test and the follow-up test having an affect on the results), as the same researcher conducted all the tests, and it is likely that they have established their own standardised routine. Again, the internal reliability of this study is enhanced by the same MC completing the tests as it would have lessened the likelihood of instrumentation threat, or something in the way the initial and follow-up assessments were done. The follow-up provided additional information for 22 of these 48 cases. For three of the children the additional information changed the outcome of the case. For three of the children the follow-up refuted tentative findings from the first assessment, however it did not change the outcome because of other physical findings. Limitations of the study included 35% of the cases who were recommended for follow-up surveys had an incomplete series, did not return for follow-up, or received only a bone scan and not a follow-up. Hence, there was incomplete data for some cases. Also, there may have been bias due to the single radiologist having an awareness of the clinical history for each case, and that there was not any multi-observer analysis of radiographic findings. Again, the small sample numbers has greatly reduced the power of the study, and necessitated the use of Chi-Square test of significance to analyse the results. As above, it would have enhanced this study to use a larger sample, to provide generalisation of results to the wider population, and so enhanced the study's external validity. 3.0 Method This chapter details how the information was gathered for this research, as well as discussing the results of data collection. Also presented are the rational for the method used including details of validity, reliability and generalizability. Secondary data can be defined as information collected by researchers that are used for purposes other than its original intention (Crombie, 1996; Crookes, & Davies, 2004; Saunders, Lewis, & Thornhill, 2005). Secondary data can be used as a research method for comparing studies on a given topic (Gerrish, & Lacey, 2006; Padget, 2004). The advantages of using secondary data are that the available information in existing published studies may be completely adequate to answer the research question of another study (Charles, 2002; LoBiondo-Wod. & Haber, 1998; Saunders, Lewis, & Thornhill, 2005). The methodology concerned performing a systematic literature review. This type of review is emphasised currently because the search for and evaluation of the studies is thorough with much rigour (Cormack, 2000; Parahoo, K 1997; Smith, 1997). However, it is described as “arduous and time consuming” by Mulrow,C.D (1994) cited in Parahoo (1997) “but quicker and less expensive than a new study”. The aim of a systematic literature review is “to provide a list as comprehensive as possible of primary studies, both published and unpublished” (Khan, et.al 2001; Beverley, Edmunds-Otter, & Booth, 2006). The intention was to locate earlier studies and reports on the issue of imaging infants diagnosed with non-accidental injury. This would assist in answering the study question: “should the follow up skeletal survey be mandatory in the assessment of non-accidental injury (NAI) in infants?” A systematic literature review was conducted instead of an empirical study, as it was unlikely that an ethics committee would consent to an empirical study being carried out that concerned children. However, the original authors of the articles would have to have permission from the ethics committee. The systematic process allows information to be gained from a wide range of sources. The studies and observations previously conducted by practised professionals were fairly accessible. This method would therefore, provide more in-depth information to enable the aims and objectives of this study to be achieved. The secondary data for this research was collected using the Internet and the university library. The university provides access to online journals through a library portal on their home web page. This link provided the database Science Direct (SD) 1966-2007 through which all the relevant articles were found. The search was instigated in December 2006 through January 2007 in order that the most recent published literature was found. The advantages of using the Internet included the ease of use of the medium, the access to current data, the speed with which data could be accessed, and the ability to source large amounts of quality data that met the research objectives. In turn, the advantage of using an online database is the ease of use and the speed at which up-to-date precise information can be sourced. The method for the search used certain tools. These tools included using keywords to refine the process whereby articles were only included if they contained the keywords as denoted by the searcher (Appendix B). 4.0 Discussion It is clear that objectives of this study were met in that the variety of literature sourced through a comprehensive literature search supported the use of skeletal survey for precise assessment, and of follow-up survey to confirm suspected cases of NAI. The factors which make follow-up skeletal survey an obvious choice to confirm suspicions of NAI include its ability to provide additional information such as fractures not identified in the initial assessment or the disconfirmation of suspected fractures from previous assessments (Block, 2005; Day et al, 2006; Kleinman et al, 1995; Williams & Treves, 2007; Zimmerman et al., 2005). Kleinmann et al. found that of the 70 fractures identified during the first skeletal survey, 13 of these had additional information imparted about their age during the follow-up skeletal survey. It was concluded that when there is suspected child abuse due to an initial skeletal survey and other examinations, that a follow-up skeletal survey be used to allow for a more thorough and comprehensive analysis of osseous injuries. Halliday et al. (2003) too emphasised that the follow-up survey was helpful in aiding clinicians to observe and document all of the fracture injuries of a specific case. Zimmerman et al (2005) highlighted that the additional fractures identified or clarified by way of follow-up survey provided the researchers with support for follow-up skeletal survey use. They also concluded that all children who have an initial assessment because of suspected NAI should have a follow-up. Especially as 27 previously unsuspected fractures were detected, the follow-up is touted as a valuable method for confirming suspected NAI. Overall, the follow-up skeletal survey has greater sensitivity in picking up metaphyseal lesions, and is much more accurate in providing information about the age of fractures consistent with the callus formation during the healing stage. The research studies presented support many others in the argument that follow up skeletal survey is essential in confirming suspected child abuse (Belfer, Klein, & Orr, 2001; Mandelstam, Cook, Fitzgerald, & Ditchfield, 2003; Williams & Treves, 2007; Ziegler, Sammut, & Piper, 2005). However almost all of the studies failed to clearly state their hypothesis and utilised scant literature reviews to develop their research design, and to justify the reasons behind the approach that they used. Using a large sample, Halliday et al. had three hundred and twenty-three hospitals with both paediatric and radiology departments that were contacted to take part in the study. The Medical Directory was used to locate the hospitals. However, a random sample of the eligible hospitals was not used, and this of course would lower its external validity, or generalizability of results to hospitals which did not take part in the study. The selection of hospitals could have been biased through the non-random selection method, for example ignoring small towns that were considered to perhaps lack the numbers of NAI cases required to see an effect in the study. This is a serious consideration, given that there was only a 30% response rate (n = 97). Also, there was no indication about the internal reliability of the questionnaire itself, and so it is assumed that the questions were mostly open ended, yes/no, true or false or some other categorical form. If Likert scale items had been used than a total survey score could have been used in correlations, ANOVA or regression analysis and so increase both the internal and external reliability and validity of the study. The other studies relied upon small sample numbers, hence non-parametric tests such as Chi Square tests of significance were used for analyses. As such, it is unlikely that samples had normal distributions and so their generalisability (external validity) to the populations from which they were drawn is limited. Apart from comparing follow-up results to initial assessments, each of the studies did not use a comparison group. The range and diversity of initial assessments would have made it difficult to compare post results with initial findings with great confidence. An internal reliability threat across all studies was that of history and maturation effects, as some samples had over two weeks between surveys, and not all participants were in Department of Children care, and so ongoing abuse may have confounded the results. Mono-method bias was high across all the studies except for Halliday et al. (2003) as alternate forms of measurement were not used, and so reduced the validity of the single method used. As well, the samples were not fully described across the studies in terms of gender, age range, ethnicity or socio-economic background, which would have increased the external validity of the studies. Overall, the most comprehensive study was by Day et al. (2006) as it specifically looked at the viability of using follow-up skeletal surveys for confirming NAI in the clinical setting in the UK, and of vouchsafing its predictive ability. The design of their study did not provide a representation of the entire population of children exposed to NAI who may have found benefit from follow-up skeletal survey; however, it did provide a snapshot of the population. Their results reflected the findings of other studies used in this paper's literature review, which confirms the critical importance of using follow-up skeletal survey to confirm identified fractures in cases of suspected child abuse. That follow-up surveys identified more rib fractures as compared to initial surveys is in agreement with other studies reviewed in this paper. Across cultures, the UK and the USA, it appears that similar results are found pertaining to the advantages of using follow-up skeletal survey. This is to be expected given that both are advanced industrialised nations which share knowledge, technology and skills by way of peer-reviewed journals, conferences and collaborative research practices ( ). Hence, it would be unlikely that particular methods of one of these nations that were found to be of great benefit in confirming NAI cases would not be taken up by the other, given their close associations both within academic and wider socio-economic spheres. However, it is clear that that the UK needs to extend research into this area, and to better ascertain the practices of follow-up radiography that are currently used. Within the USA national guidelines for follow-up skeletal survey are already in place, and this has aided standardisation of practices and of research endeavours, allowing for potentially larger samples and better comparison groups, given that the same procedures are throughout each state. This is an important limitation in the UK at present, and could explain the small sample numbers for research studies given the variety of follow-up radiographic techniques to choose from, and the diversity of processes. At present there are only a few predictive criteria which aid the clinician in identifying which children to recommend for follow-up imaging procedures, and the skeletal survey is one such method which can aid in this area. None of the papers used in the literature review of this paper considered follow-up skeletal survey to be of limited or negligible use, in fact all the evidence points to skeletal survey as being one of the most comprehensive methods. However, only two of the studies included here identified other radiographic methods to support findings of the skeletal survey (James et al, 2003; Zimmerman et al, 2005). As well, the follow-up skeletal survey offers a comprehensive method of confirming or disconfirming suspicions of NAI. The studies have culminated in the data required to answer the research question: Should the follow-up skeletal survey be mandatory in all cases of suspected child abuse in infants? Limitations and Recommendations There are however notable disadvantages of using secondary data as a research collection method. Firstly, the data tends to have been originally collected to answer a different research question (Padget, 2004). Secondly, as time passes the data may be obsolete or at least less meaningful. Also, the definitions used in the original study may not meet the requirements for which the secondary data is needed (Saunders, Lewis, & Thornhill, 2005).In regards to measurement error, such as standard deviation and standard error, not all sources of original qualitative data report these, and so the degree of precision of compiled data cannot be guaranteed (Saunders, Lewis, & Thornhill, 2005). Additionally, reliability, validity and generalizability of secondary data are low, as different researchers are likely to uncover different documentation at different times. Face validity can be checked by comparing the findings reported with the objectives set to see if they meet those objectives; however this is a highly subjective process. A disadvantage to using the Internet is that a number of databases require subscription in order to access the journals, and these costs are not always feasible. For this reason the university’s research databases were used, accessed via the library. Also, the use of inefficient key words and phrases can slow the search, or result in no data found. An important disadvantage to using the Internet is that there is low reliability, validity and generalizability of results, as over time key words changes in the degree of importance, and it is unlikely that another researcher will obtain the exact results even if they use the same web browser and search engines. Conclusion Child abuse is usually difficult to recognise and to assess accurately. Non-accidental injury needs to be identified so that physicians can alert the proper authorities as to the possibility of abuse and to apply instruments that will help to identify cases of abuse and to confirm or disconfirm suspicions. These studies paper have shown the viability of using the follow up skeletal survey as a method to confirm suspicions of NAI in children and toddlers. Skeletal injuries that are related to an inconsistent history can be assessed for the recognisable patterns which tend to be strong indicators of child abuse. Given the potential for new information to be gained from follow-up skeletal surveys it is advisable that they be used to confirm or disconfirm suspected NAI cases. However, it must also be kept in mind that contemporary research studies are still relying on small sample numbers and the use of Chi Square tests of significance to determine if a real difference exists between initial skeletal surveys and follow-up skeletal surveys. As indicated in the Literature Review and Discussion the use of small sample necessitates the need to use non-parametric tests, which have low or nil generalisability (external validity) and low reliability, given that the samples cannot confidently be considered true reflections of larger populations. For this reason it has been recommended by this study that the power of future research be increased by using larger samples, and random allocations to two or more groups, including a control group, so as to increase the confidence that any differences that exist will be detected, and that these differences are real, and not simply due to chance. It is clear that follow-up skeletal survey can indicate evidence of care and protection and can contribute critical information to criminal proceedings and other forms of litigation. Given the variation of procedures for follow-up assessment used across the UK it is essential that a standardised procedure be used to aid data collection, evaluation and comparison of suspected abuse cases. At present no national guidelines exist. Overall, the follow-up skeletal survey provides a more precise and comprehensive assessment of osseous injury. Additionally, the skeletal survey is available in the radiology departments at most hospitals and so would facilitate the standardisation of radiograph procedures across the nation as suggested by the BSPR guidelines. This would extend research efforts as procedures could be more accurately implemented, evaluated and compared. The implications of this research are the extension of understanding of the potential benefits of follow-up skeletal surveys, and indications for future research to further knowledge in this area. References Belfer, R. A., Klein, B. L., & Orr, L. (2001) Use of the skeletal survey in the evaluation of child maltreatment. American Journal of Emergency Medicine, 19(2): 122-124. Beverley, C., Edmunds-Otter, M., & Booth, A. (2006) Systematic reviews and secondary research (Chapter 21, pp. 316-334). In Gerrish, K. & Lacey, A. (2006) (Eds). The Research Process in Nursing, 5th Edition. Oxford: Blackwell Publishing. Block, R. W. (2005) Follow-up skeletal surveys prove to be valuable in evaluation of child physical abuse. Child Abuse Neglect, 29(10): 1073-1074. British Society for Paediatric Radiology [BSPR] (2003) Standard for skeletal surveys in suspected non-accidental injury (NAI) in children.[online] Available at http://www.bspr.org.uk/ (accessed 19/03/07) Bryman, A. (2004) Social Research Methods, 2nd ed. Oxford: Oxford University Press. Charles, D. (2002). Handbook of Research Design. Sagebush: Thompson Publications. Coffey, C., Haley, K., Hayes, J., & Groner, J. I. (2005) The risk of child abuse in infants and toddlers with lower extremity injuries. Journal of Paediatric Surgery, 40(1): 120-133. Cormack, D. (2000) The Research Process in Nursing, 4th Edition. Oxford: Blackwell Publishing. Crombie, I. K. (1996) The Pocket Guide to Critical Appraisal: A handbook for health care professionals. London: British Medical Journal. Crookes, P. A. & Davies, S. (Eds) (2004) Research into Practice: Essential skills for reading and applying research in nursing and health care. 2nd Edition. London: Bailliere Tindall. Day, F., Clegg, S., McPhillips, M., & Mok, J. (2006) A retrospective case series of skeletal surveys in children with suspected non-accidental injury. Journal of Clinical Forensic Medicine, 13(2): 55-59. Ellerstein, N. S., Norris, K. J. (1984) Value of radiologic skeletal survey in assessment of abused children. Paediatrics, 74(6): 1075-1078. Gerrish, K. & Lacey, A. (Eds) (2006) The Research Process in Nursing, 5th Edition. Oxford: Blackwell Publishing. James, S. L., Halliday, K., Somers, J., & Broderick, N. (2003) A Survey of Non-accidental Injury Imaging in England, Scotland and Wales. Clinical Radiology, 58(9): 696-701. Kemp, A. M., Butler, A., Morris, S., Mann, M., Kemp, K. W., Sibert, J. R., & Maguire, S. (2006) Which radiological investigations should be performed to identify fractures in suspected child abuse? Clinical Radiology, 61(9): 723-736. Kleinman, P. L., Kleinman, P. K., & Savageau, J. A. (2004) Suspected infant abuse: Radiographic skeletal survey practices in paediatric health care facilities. Radiology, 233(2): 477-485. Kleinman, P.K, Marks, S. C, Richmond, J. M., & Blackbourne, B. D. (1995) Inflicted skeletal injury: A post-mortem radiologic-histopathologic study in 31 infants. American Journal of Radiology, 165: 647-650. Kleinman, P. L., Nimkin, K., Spevak, M. R., Rayder, S. M., Madansky. M. L., Shelton, Y. R., Patterson, M. M. (1995) Follow-up skeletal surveys is suspected child abuse. AJR, 167: 893-896. LoBiondo-Wod, G. & Haber, J. (Eds) (1998) Nursing Research: Methods, critical appraisal and utilization, 4th Edition. London : Mosby. Mack, N, et al. (2005). Qualitative Research Methods: A data collector’s field guide. North Carolina: Family Health International. Mandelstam, S. A., Cook, D., Fitzgerald, M., & Ditchfield, M. R. (2003) Complementary use of radiological skeletal survey and bone scintigraphy in detection of bony injuries in suspected child abuse. Archive of Disability in Children, 88(5): 387- 390. Mason, J. (1997) Qualitative Researching: An introductory text. New York: Sage Publications. Mckinney, A., Lane, G., & Hickey, F. (2004) Detection of non-accidental injuries presenting at emergency departments. Emergency Medicine Journal, 21: 562-564. Nimkin, K. Spevak. M. R., & Kleinman, P. K. (1997) Fractures of the hands and feet in child abuse: imaging and pathologic features. Radiology, 203(1): 233-236. Padget, D. K. (2004) The Qualitative Research Experience. Southbank: Thompson Learning. Parahoo, K. (1997) Nursing Research: Principles, process and issues. London: Palgrave Macmillan. Saunders, J., Lewis, S., & Thornhill, M. (2005) Research Methods for Medical Students. New York: Sage Publications. Shank, G. D. (2002) Qualitative Research: A personal skills approach. Columbus: Merrill/Prentice Hall. Smith, P. (Ed) (1997) Research Mindedness for Practice: An interactive approach for nursing and healthcare staff. London: Churchill Livingstone Williams, S. & Treves, J. (2007) A second radiographic skeletal survey for child abuse triggered by bone scintigraphy found positive after the initial survey was called negative. Clinical Nuclei Medicine, 32(1): 29-31. Ziegler, D. S., Sammut, J., & Piper, A. C, (2005) Assessment and follow-up of suspected child abuse in preschool children with fractures seen in a general hospital emergency department. Journal of Paediatric Child Health, 41(5-6): 251-255. Zimmerman, S., Markoroff, K., Care, M., Thomas, A., & Shapiro, R. (2005) Utility of follow-up skeletal surveys in suspected child physical abuse evaluations. Child Abuse & Neglect, 29(10): 1075-1083. Appendix A Table 1 Complete Skeletal Survey Appendicular Skeleton Humeri (AP) Forearms (AP) Hands (PA) Femurs (AP) Lower legs (AP) Feet (PA) or (AP) Table 2 Follow-up procedures used. Table 3. Occult injuries detected. Fracture % Rib fractures 52 Tibial 19 Femoral 9 Figure 1. Skeletal survey and follow-up skeletal survey of a 14 month-old abused boy. A. Initial skeletal survey, lateral radiograph of sacrum, reveals well-marginated bone density as indicated by the arrow, this suggests a fracture. B. Follow-up survey a fortnight later shows union of fragments, as indicated by the arrow, and confirms that the injury was recent when A was obtained (Kleinman et al, 1995, p. 895). Appendix B The keywords used were non-accidental injury, child abuse, imaging, protocols, children and follow up skeletal surveys. Six comprehensive searches were carried out first, after that all searches came back with the same articles. Search one consisted of the keywords ‘non-accidental injury’ in the Science Direct database. In search two, the key words were ‘imaging’ and ‘child abuse’. Search three consisted of the keywords ‘imaging’ and ‘child abuse’ and ‘protocols’. For search four the keywords were ‘protocols’ and ‘non-accidental injury’ and ‘imaging’. Search five was made through the ISI Web of Knowledge with the keywords ‘non-accidental injury’ and ‘children’ and ‘imaging protocols’. Search six used the same database and keywords as before with the addition of one more key word ‘skeletal survey’. Some of the articles were cross referenced to find further relevant articles. Key words used in literature search skeletal survey follow-up skeletal survey radiographic imaging non-accidental injury (NAI) child abuse +protocols +imaging +children +infants +toddlers Read More