This assignment will critically appraise a research article titled ‘The association of responsiveness in oral and general health-related quality of life with patients’ satisfaction of new complete dentures’, published online on 22nd November 2012.AbstractThe abstract of a study is a summary of the content of the journal manuscript (www.springer.com). In this article, information regarding the study is simply yet not entirely summarised in the abstract. However, to some degree, this would emphatically save researchers’ time in determining the fitness of the article as their reference.Seen in the first section is the purpose of the study. The purpose should determine the scope, depth and the overall direction of the research (Research Methodology) including the immediate and any larger, eventual purpose of the research (Unk.edu). In this section, the study purpose can be seen divided into two parts: “to determine the responsiveness properties of the Oral Health Impact Profile (OHIP) short-forms and 36-item Short-Form (SF-36) in complete denture treatment” and “to evaluate the association between patients’ satisfaction and improvements in oral and general health-related quality of life (OHRQoL and HRQoL) after fitting of new, complete dentures”. However, no interrelation between these two purposes is seen when in fact, the word ‘further’ suggests that the second purpose can only be done following the results from the first one. Although, the interrelation is mentioned later that OHIP and SF-36 are used as OHRQoL and HRQoL measures respectively, their sudden appearance in the beginning of the abstract is questionable to readers as they are not mentioned in the title. On top of that, ‘discriminatory ability’ is not mentioned here yet seen in the introduction in ‘aim’ paragraph.Fowkes and Fulton (1991) mentioned that study design should be stated in the abstract at the beginning of the method section but this is not seen in the article making readers unsure of what aspects of appraisal to concentrate on. Study designs articulate what data is required, what methods are going to be used to collect and analyse this data (Wyk, n.d.). Hence, crucial before executing a research. It is, in turn, paramount for the subsequent quality of the research, the reliability of its conclusions, and the ability to publish a study (Altman et al, 1983) yet this significance is underestimated by the researchers. Examples of a study design are case report, cross-sectional study and cohort study (Fowkes and Fulton, 1991). Although ‘cohort studies’ is stated in the introduction, it is not specifically used to denote the study as one. Nevertheless, it can be deduced that it is a cohort study, taking into account its general implementation that matches ‘cohort study’ definition. A cohort study is an observational study of a group of people with specified characteristics who are followed up over a period of time to detect new events (Fowkes and Fulton, 1991). Despite this issue, research site, the number of subjects, the overview of the subjects’ intervention and the analysis methods used are presented concisely.Both of the results and conclusions in the abstract too are summarised concisely. Only statistically significant results corresponding to each of the study aims are presented. Moreover, the conclusions are aptly inferred based on the results and all other information described earlier in the abstract.KeywordsFollowing the abstract are keywords that should represent the content of the manuscript (www.springer.com). Nothing matters in this section because they are relevant and do represent the whole article.AbbreviationsWith numerous applications of long technical terms in this article such as oral and general health-related quality of life, Oral Health Impact Profile and 36-item Short Form, it is certainly a great idea to provide the abbreviation section. This section outlines all the long terms together with what they stand for. Alongside the fact that the terms are not abbreviated when first used, this greatly facilitates readers to go through the lengthy article as they can just simply refer to this section should they forget what they stand for. IntroductionThe researchers commence the introduction by mentioning the importance of patients’ perceptions in dental care. Readers can be assured on the use of patients as measures for quality of life because as declared, this is just a replication of method used by four other researchers previously (Saito et al, 2010; Naito et al, 2010; Berretin-Felix, 2008; John et al, 2004). They continue by telling readers how OHRQoL and HRQoL have been popularly measured by OHIP and SF-36 respectively. Their utilisation validity is supported by a study stating that SF-36 are most commonly used for assessing HRQoL (Lu et al, 2003) and another reporting that OHIP-49 serves better to measure the OHRQoL than the SF-36 (Allen et al, 1999). Similar to OHIP, SF-36 has also been translated into Taiwanese (Chinese) version. Although one issue arose that the translated version is somewhat adapted to Western instead of Chinese cultures, Tseng et al (2003) found that it is in fact, a practical and reliable instrument in the general population. Along with studies by Kuo et al (2009) and Lee et al (2007) as stated in the article, this further strengthens the translated SF-36 utilisation validity in measuring HRQoL.Next, the researchers state that the investigation of patients’ satisfaction with dentures in relation to their quality of life still remains an interest. Readers would now know why this is made as one of the aims of the study.Materials and MethodsWith reference to the ‘cohort study’ definition, uniform subjects’ characteristics are required for the study. In the study, subjects are edentulous elderly aged 65 years and over with low familial income. According to the United Kingdom Adult Dental Health Survey (2009), more than 15 percent of adults aged 65 years and over are edentulous and require complete dentures (Table 1). This is further supported by a data from The United States National Health and Nutrition Examination Survey (NHANES) that shows an overall percentage of 27.27 percent of seniors over 65 years old with total tooth loss (Table 2) while only 3.75 percent for those within 20 to 64 years old (Table 3). Also considered in the latter survey was their poverty status where higher poverty status is indicated by lower familial income. Results show that those with high poverty status are more likely to be edentulous. On a side note, by having only the low-incomers in the welfare plan, bias is avoided because high-incomers should be able to afford the dentures since dentures are given without charge. The notable disparity of mentioned characteristics between different age groups confirms subjects’ dentitions to be edentate. This shows that characteristics set was ideally correct and correspond with the aim of the study relating to complete dentures. CharacteristicsPercentage with no remaining teeth (%)Age65 to 74 years1575 to 84 years3085 years and more47Table 1. Data regarding tooth loss in adults aged 65 years and over from The United Kingdom Adult Dental Health SurveyCharacteristicsPercentage with no remaining teeth (%)Age65 to 75 years23.9375 years or more31.30Poverty StatusLess than 100%44.19100% to 199%36.61Greater than 200%17.25Total27.27Table 2. Data regarding tooth loss in adults aged 65 years and over of different poverty status from The United States National Health and Nutrition Examination SurveyCharacteristicsPercentage with no remaining teeth (%)Age20 to 34 yearsNot enough data35 to 49 years2.6350 to 64 years10.13Poverty StatusLess than 100%9.28100% to 199%6.48Greater than 200%2.35Total3.75Table 3. Data regarding tooth loss in adults aged 20 to 64 years of poverty status from The United States National Health and Nutrition Examination SurveyAlso stated here is that only the funding for manufacturers’ services is limited but not for the denture procurement. This suggests that denture processes are not controlled, for example, manufacturers are free to decide denture materials. However, in the discussion section, the researchers also state that the materials for dentures were standardised for all patients so these contradicting statements may confuse readers about the uniformity of denture materials. Therefore, the ‘procurement’ should have been clearly elaborated in this section.Informed consent is an inevitable requirement prior to every research involving human as subjects (Nijhawan et al, 2013). In this study, the selection procedure chosen is described very vaguely and not a clue about this crucial prerequisite is documented. There is a consensus that this is immoral because it is important to procure subject’s willingness after consenting based on the information given (Nijhawan et al, 2013). In this case however, due to age factor, the seniors would potentially sign the consents without being fully aware of what they are signing. This is feared to result in the withdrawal of these clueless subjects eventually (Nijhawan et al, 2013). Additionally, the total number of patients under the plan is not reported so it is difficult for readers to judge the sample size validity of the study in relation to the welfare plan population.Despite the lack of comprehensive selection approach, the fact that the study was approved by an Institutional Review Board (IRB) preponderates over the mentioned appraisal. The primary purpose of IRB is to protect the rights and welfare of human subjects involved in research activities being conducted under its authority (University of Pittsburgh). For this reason, correct research protocols must have been fulfilled. Informed consents, for instance, would have been issued to patients in a language which is easily understood by them (Nijhawan et al, 2013). However, it appears that the researchers are confident that readers would understand how the study design is perfectly valid just by providing the ethical approval statement. This is one prominent weakness as not all readers are familiar with IRB roles and their standards to meet their approval. The researchers carry on to tell that 5-point scales are used to assess patients’ satisfaction and for OHIP-49 responses (Likert scales). Adding to the fact that people define ‘well-being’ differently, the scales could give subjects difficulty to decide as to which level they are at. Patients would feel easier if more options of ‘satisfaction’ and ‘experience’ levels are provided. Furthermore, subjects are required to answer all OHIP versions besides the OHIP-49 and SF-36 questionnaires so it is very time-consuming for them.The researchers then describe methods to statistically analyse each of the aims clearly. For some methods, for example, Cohen’s ES, they even tell readers how they are calculated. Furthermore, readers can be convinced that only minute errors could have happened during the analyses because they are done electronically with SAS9.1.3. However, the validity of the analysis methods’ utilisations is uncertain as there is no information supporting this.ResultsGenerally, all results are neatly tabulated with descriptions and those which are statistically significant are bold as in tables 1, 2 and 3. Thus, are very reader-friendly. In the first paragraph, it is declared that each participant completed both pre- and post-treatment questionnaires but no data about them remaining throughout the study is shown to confirm the statement. The need to declare this is because Song and Chung (2010) reported that prospective—done from the present time into the future—studies are vulnerable to a high loss to follow-up rate due to subjects’ withdrawal at later stage of the study. Not to mention, the 6-month follow-up period is long enough for this to potentially happen. Nonetheless, the duration set is sufficient to discern the significant (but not optimum) effects of denture treatments had on the subjects meaning results are acceptable.In table 1, the internal consistency of all OHIP versions are at least 0.89 which confirms their reliability as measures in the study. Moreover, the use Scheffe post hoc comparisons enables readers to understand the results of perceived oral health more easily because they are converted into words. However, no reason is provided as to why SF-36 results are not included in the table. In table 2, the number of subjects for different demographic characteristics is indicated by ‘n’. One prominent error is that for ‘IADL’ variable, the total ‘n’ is less than 224. This suggests that some of the patients’ responses are either accidentally or purposely omitted. If this was done purposely, the reason for the omission should have been declared. Furthermore, more than half of the subjects have at least one chronic disease and abnormal IADL, suggesting that most of them have carers. Although clinical knowledge is not required prior to interview because there is no clinical terms in the questionnaires, it is possible that responses are from the carers instead of the patients themselves. Thus, only based on the carers’ knowledge gained when looking after the patients. A study from University of Sheffield reported that research that seeks to establish the circumstances, preferences and views of elderly by asking carers is bad science and unethical. In addition, Hawthorne Effect may occur where subjects’ behavior is altered due to the presence of interviewer, providing erroneous responses (Spicker, 2007). As a consequence, results might have been inaccurate, making the entire study less reliable. Adding to that, little information regarding patient interviews is reported. It is unspecified whether the interviews are done in groups or individually. Due to the association of patients’ well-being in the questionnaires, this information should have been included because their responses could vary in between group and individual interviews. In group interviews, their responses could be affected by others (Kitzinger, 1994a, b; De Jong and Schellens, 1998). This is in contrast to individual interviews as patients are more open with sensitive-topic discussions (Kaplowitz, 2000) so their in-depth information can be collected at a time and then compared (Oppenheim, 1992; Weiss, 1994; Knodel, 1997). Therefore, it is feared that the interviews may have been done inconsistently—some in groups and others individually—because the differences in information collected may rather be a function of their behaviours in group, not an indicator of the data validity (Kitzinger, 1994).Although compact, all data is well-organised in table 3. Results in table 4 shows that patients only have medium satisfactions with their new dentures after 6-month treatments. If treatment duration was extended, denture effects would be more apparent and these results would increase to above ‘medium’ level. In table 5, the relationship between patients’ satisfaction with new dentures and quality of life is seen measured only using SF-36, OHIP-49 and its version with the best discriminatory ability, OHIP-14T. This is maybe in order to observe only the optimum results.DiscussionThe discussion section should clearly link the research to the studies reviewed earlier, compare the new findings to and discuss how the study has reinforced previous findings and state the need of further research. Firstly, the researchers refreshed readers with the summary of the results from tables 1, 3 and 5. Discussions only begin in the second paragraph where the discriminatory ability and reliability of all OHIP versions results are compared with those from previous studies (Allen and Locker, 2002; Ellis et al, 2007). The previous ones only either assessed the discriminatory ability or internal consistency to determine the measure utilisation suitability but Kuo et al assessed both features for each of the versions. Only in terms of internal consistency, Kuo et al confirm the previous studies’ results but not of discriminatory ability. It is shown that the present and previous findings are close suggesting that patients’ perceptions of oral health vary from different regions as the previous studies were conducted in Brazil and UK.After that, readers are assured of the validity in using Cohen’s ES to identify differences as four previous studies (Saito et al, 2010; Berretin-Felix et al, 2008; Allen and McMillan, 2003; Wong et al, 2007) have used the same method. Although a slight drop is seen in the ES of OHIP-49 when compared to Allen and McMillan’s finding, it still falls in the same category (small) with the previous one. The slight difference may be resulted from the difference in subjects’ characteristics where CDG2 group is patients receiving denture replacements only while in the present study, subjects are mixed with first-time denture wearers. With regard to OHIP versions, OHIP-14T and OHIP-EDENT have the highest ES in the present and previous study respectively. However, two identical dimensions: ‘physical pain’ and ‘psychological discomfort’ of both the different versions are found to have significant improvements meaning similarity is only observed in the dimensions but not the whole version.Then, the researchers reveal that their finding on EF of SF-36 is small and similar to Heydecke et al’s. Even the highest EF they observed is ‘GH’ dimension and is still below the range of ‘small’ ES. Thus, confirming the previous finding. Moreover, based on the overall EF results, they discover that OHIP is more sensitive than SF-36.Next, the researchers relate their findings about patients’ satisfaction in OHRQoL with those of Stober et al’s as consistent. However, the depth of the consistency is uncertain because both studies used different measures. Furthermore, the researchers admit that due to the low sensitivity of SF-36, patients’ satisfaction is not significantly associated with HRQoL.The researchers then proceed by recognising the study limitations along with the strength. Due to the absence of control group in this study, they believe future investigation is required with high-incomers. On the other hand, no discrepancies are encountered as they review clinic services regularly to ensure denture materials used are uniform.ConclusionWith minute contrasts in results when compared to previous studies, all OHIP versions and SF-36 are still favourably valid to measure quality of life. It is only the study design that needs improvement. Readers can be more confident if it was described thoroughly particularly the subject selection approach because the ethical approval alone does not give enough clarification to them. Thus, welcoming fear of results being inaccurate to readers. Otherwise, all statistical analysis results should be accurate because they are calculated electronically. Furthermore, the investigation of patients’ satisfaction in relation to the quality of life has now been accomplished, providing new information into the healthcare but still, future research is needed. Therefore to conclude, Kuo et al’s findings can surely aid future research to large extent. TitleWhen the article title is read for the first time, it is not immediately grasped because it is quite long. The title lacks the specification of the patients—of which area or nationality—so readers are unable to fathom where the research revolves around. Although this, with the presence of major key points such as ‘quality of life’, ‘patients’ satisfaction’ and ‘complete dentures’, the complicated title actually condenses the whole article (Bavdekar, 2016) and encapsulates its summary to large extent. This shows that the title is exhibited concisely in an imprecise manner with substandard clarity. Thus, conveying only most of the ideas to readers on what to predict in the article but not enabling them to envision it wholly. AcknowledgementOne flaw noticed in this section is that there should have been more parties recognised apart from the Health Bureau of Kaohsiung County Government (ROC) Taiwan for partly granting this study. Without the subjects’ involvements, the study could not have been performed so they should have been acknowledged for their willingness to participant in the study. This goes similarly to denture manufacturers for producing massive 224 dentures for patients. Thus, insinuating the researchers’ lack of consideration towards those who contributed into the study.AuthorsThe researchers’ contributions towards the study are believed to be unbiased because this has been verified in the acknowledgement section. A consensus must have been fulfilled between them when delegating roles for the study before its commencement. Furthermore, it is disclosed in the article that the researchers—except for H.-C. Kuo and J.-C. Wang—have different expertise in the healthcare. Y.-S. Kuo from the Health Promotion Division for example, could have been the one searching for potential subjects as the welfare plan could be established under this department. Additionally, it would not have been feasible for only one person to follow-up 224 patients in just a 6-month period. Therefore, both H.-C. Kuo and J.-C. Wang from College of Dental Medicine, believed to be dental practitioners are involved most likely in terms of monitoring the subjects. Not to mention, statistical analysis would have been led by the one and only researcher majoring in this field, Y.-H. Yang from the Division of Statistical Analysis. The rest of them would have contributed with regards to patient interviews, evaluation of patients’ satisfaction, data compilation and publishing of the article.