Statistical significance, clinical importance and effect sizes: Enhancing understanding of a study's results.

BACKGROUND: The proper interpretation of a study's results requires both excellent understanding of good methodological practices and deep knowledge of prior results, aided by the availability of effect sizes. METHODS: This review takes the form of an expository essay exploring the complex and nuanced relationships among statistical significance, clinical importance, and effect sizes. RESULTS: Careful attention to study design and methodology will increase the likelihood of obtaining statistical significance and may enhance the ability of investigators/readers to accurately interpret results. Measures of effect size show how well the variables used in a study account for/explain the variability in the data. Studies reporting strong effects may have greater practical value/utility than studies reporting weak effects. Effect sizes need to be interpreted in context. Verbal summary characterizations of effect sizes (e.g., "weak", "strong") are fundamentally flawed and can lead to inappropriate characterization of results. Common language effect size (CLES) indicators are a relatively new approach to effect sizes that may offer a more accessible interpretation of results that can benefit providers, patients, and the public at large. CONCLUSIONS: It is important to convey research findings in ways that are clear to both the research community and to the public. At a minimum, this requires inclusion of standard effect size data in research reports. Proper selection of measures and careful design of studies are foundational to the interpretation of a study's results. The ability to draw useful conclusions from a study is increased when investigators enhance the methodological quality of their work.

The Sinhala version of the pain catastrophizing scale: validation and establishment of the factor structure in pain patients and healthy adults.

OBJECTIVE: This study was conducted to translate the Pain Catastrophizing Scale into and adapt it to the Sinhala language and to examine its psychometric properties and factor structure in pain patients and healthy adults in Sri Lanka. SETTING AND DESIGN: A cross-sectional study was conducted, recruiting pain patients from multiple clinics and healthy adults from the community as convenience samples. METHODS: Cross-cultural adaptation of the Pain Catastrophizing Scale for Sinhala speakers was carried out using recommended methods. The adaptation's psychometric properties and factor structure were tested in 149 pain patients and 172 healthy adults. Temporal stability was tested in a sample of 104 young adults. Pain intensity of patients was assessed using a visual analog scale, and personality traits of all participants were assessed with the Eysenck Personality Questionnaire. RESULTS: Factor analysis revealed that the three-factor structure of the original version of the Pain Catastrophizing Scale was the best fit to the data from participant samples. Cronbach's alpha values of the three components and total scores for patients and healthy adults ranged from 0.72 to 0.87. Pain catastrophizing exhibited moderate positive correlations with neuroticism in patients and healthy adults and with pain intensities in patients. A high intraclass correlation coefficient of agreement (0.81) revealed an acceptable temporal stability in young adults. CONCLUSIONS: The results suggest that the Sinhala version of the Pain Catastrophizing Scale retains the original three-factor structure. It is a stable, valid and sufficiently reliable tool to assess pain catastrophizing in Sinhala-speaking individuals in Sri Lanka.

Headache and oral parafunctional behaviors.

This study tested the hypotheses that individuals with headaches would show significantly more oral parafunctional behaviors than non-headache controls, be diagnosed with one or more temporomandibular disorders (TMD) significantly more frequently than controls, and would report significantly less pain and other symptoms of headache after participating in a habit reversal treatment to reduce oral parafunctional behaviors, compared to a wait list control. In Phase I, individuals with and without self-reported headaches were examined by a blinded examiner and participated in a week-long experience sampling protocol (ESM) to assess oral parafunctional behaviors, pain, and emotional states. In Phase II, those with headaches were randomly assigned to either a habit reversal treatment or to a wait list control group. In the last, sixth week of the program, participants again completed an ESM protocol. Results showed that headache patients were significantly more likely to report oral parafunctional behaviors than non-headache controls and to receive a Research Diagnostic Criteria/TMD diagnosis. Results from Phase II showed general improvement in both groups on pain and parafunctions. Individuals with headaches engage in significantly higher rates and intensities of oral parafunctional behaviors. Treatment of these behaviors using habit reversal techniques appears to have the same effect on pain as waiting.

Autonomic dysregulation in headache patients.

To analyze autonomic nervous system activity in headache subjects, measurements of heart rate variability (HRV), skin temperature, skin conductance, and respiration were compared to a matched control group. HRV data were recorded in time and frequency domains. Subjects also completed self-report questionnaires assessing psychological distress, fatigue, and sleep dysfunction. Twenty-one headache and nineteen control subjects participated. In the time domain, the number of consecutive R-to-R intervals that varied by more than 50 ms and the standard deviation of the normalized R-to-R intervals, both indices of parasympathetic nervous system activity, were significantly lower in the headache group than the control group. Groups did not differ statistically on HRV measures in the frequency domain. Self-report measures showed significantly increased somatization, hostility, anxiety, symptom distress, fatigue, and sleep problems in the headache group. The results suggest headache subjects have increased sympathetic nervous system activity and decreased parasympathetic activity compared to non-headache control subjects. Headaches subjects also showed greater emotional distress, fatigue, and sleep problems. The results indicate an association between headaches and cardiovascular functioning suggestive of sympathetic nervous system activation in this sample of mixed migraine and tension-type headache sufferers.

Tooth contact versus clenching: oral parafunctions and facial pain.

AIMS: To test the hypothesis that estimates of time spent in tooth contact are significantly greater than estimates of time spent clenching, and to test the hypothesis that tooth contact is greater in pain patients, particularly those reporting facial or head pain, than those with pain elsewhere in the body. METHODS: An anonymous, voluntary, confidential questionnaire was administered to 235 patients seeking care at a general medical clinic. The questionnaire assessed demographic variables, presence and location of pain, and percentage of time spent in tooth contact and in clenching. Analysis of variance was used to examine differences among groups of patients; logistic regression was used to identify significant predictors of pain. RESULTS: All patients reported that the percentage of time spent in tooth contact was significantly greater than the time spent clenching. The same pattern of results emerged for those with and without head pain, and those with and without any chronic pain problem. Both tooth contact and clenching were significantly associated with head pain. CONCLUSION: Results from the logistic regressions provide convergent validity on the importance of oral parafunctions, specifically tooth contact and clenching, to facial/head pain. For assessment of oral parafunctional behaviors, inquiries that utilize clear behavioral referents (tooth contact versus clenching) are likely to result in more accurate estimates than behaviors with unclear definitions.

Diurnal variation in pain reports in temporomandibular disorder patients and control subjects.

AIMS: To test the hypothesis that temporomandibular disorder (TMD) patients have characteristic diurnal patterns of pain that are associated with diurnal or nocturnal parafunctions. METHODS: Experience sampling methods were used to obtain information on pain from subjects (n = 84) diagnosed, according to the Research Diagnostic Criteria for TMD, with myofascial pain, myofascial pain and arthralgia, disc displacement, and from non-TMD controls. Variations in pain as reported on the pager questionnaire form were modeled as linear, exponential, and quadratic effects. RESULTS: Between 8.7% and 23.8% of TMD subjects with pain showed significant patterns to their daily pain reports, compared to 4.5% of non-TMD controls. Groups did not differ significantly in the proportions of those with increasing (59.5%) vs. decreasing (40.5%) pain levels. Self-reported clenching during the day and grinding at night were weakly associated with an increasing or decreasing pattern of pain during the day (P < .10). Pain levels during weekends were significantly lower for all groups. CONCLUSIONS: Strongly linear or curvilinear patterns of pain were not characteristic of this sample of subjects. More than half the subjects reported slightly increasing pain during the day, but the variability within groups was considerable. Increasing and decreasing patterns of pain were independent of self-reported daytime and nighttime clenching and grinding. Self-reported pain patterns may not be used to reliably infer the times when parafunctional activities occur. The presence of lower pain levels during the weekend probably reflects reduction in psychosocial stressors associated with the work week.

Temporomandibular disorders and facial pain: a psychophysiological perspective.

This article presents a psychophysiological perspective on temporomandibular muscle and joint disorders (TMJD) and facial pain. After a brief introduction to TMJD, the article presents data, largely derived from work carried out in my laboratory, that address four questions: (1) What are the consequences of parafunctional activities? (2) Do TMJD patients engage in parafunctional activities? (3) Why are TMJD patients unaware of these activities? and (4) What are the implications of these findings for treatment? The findings suggest that low-level parafunctions increase pain in otherwise pain-free individuals and can produce symptoms sufficiently severe to meet the diagnostic criteria for TMJD diagnoses of myofascial pain and/or arthralgia. Patients with certain forms of TMJD report very high levels of parafunctional tooth contact. Their lack of awareness of these behaviors may arise from uncertain definitions of the term "clenching", from proprioceptive deficits, or from the presence of adjunctive behaviors. Preliminary work shows that reduction in tooth contact via habit reversal techniques may be a promising mechanism for reducing pain in these patients.

Properties of silorane-based dental resins and composites containing a stress-reducing monomer.

OBJECTIVE: To evaluate properties of silorane-based resins and composites containing a stress reducing monomer. METHODS: Resin mixtures and composites were formulated containing (a) a developmental stress reducing monomer [TOSU; Midwest Research Institute]; (b) Sil-Mix (3M-ESPE); (c) photo cationic initiator system. Standard BISGMA/TEGDMA resin (B/T) and composite (Filtek Z250) were used as controls. Polymerization volume change was measured using a NIST mercury dilatometer and polymerization stress using an Enduratec mechanical testing machine. Three point bend tests determined flexural elastic modulus, work of fracture, and ultimate strength (ADA 27; ISO 4049). Fracture toughness was measured using ASTM E399-90. Four groups of resins and composites were tested: Sil-Mix, methacrylate standard, and Sil-Mix with two addition levels of TOSU. An ANOVA was used and significant differences ranked using Student-Newman-Keuls test (alpha=0.05). RESULTS: Polymerization stress values for resins containing TOSU were significantly less than the other materials. Polymerization shrinkage values for Sil-Mix formulations were significantly less than for B/T, but were not different from each other. TOSU-containing formulations generally had somewhat lower mechanical properties values than Sil-Mix or B/T. Polymerization stress values for Sil-Mix-based composites were significantly less as compared to Z250. The 1wt.% TOSU composite had the lowest stress. No difference between composite groups was noted for fracture toughness or work of fracture. For ultimate strength, the 5wt.% TOSU formulation differed significantly from Z250. All Sil-Mix formulations had elastic modulus values significantly different from Z250. SIGNIFICANCE: The ability of TOSU to reduce polymerization stress without a proportional reduction in mechanical properties provides a basis for improvement of silorane-based composites.

Psychophysiological definitions of clenching.

This study tested the hypothesis that individuals show considerable variability in EMG activity produced by the masticatory muscles when they are instructed to clench than when they are instructed to make minimal or maximal contact. Twenty individuals without temporomandibular disorder (TMD) pain participated in a biofeedback-training task to establish a relaxed baseline. They were instructed to clench their teeth according to their personal definition of the term, while EMG data were collected. This process was repeated two more times, followed by similar instructions to make minimal and maximal contact between the teeth. Results showed that individual subjects were very consistent in their behavioral definition of clenching and that the subjects taken as a whole showed markedly greater variability. The precise behavioral meaning of clenching varies across individuals. The failure to account for these individual differences may explain in part reported discrepancies on the role of parafunctions in TMD.

A comparison between two types of radiographic film for accuracy of measurements of approximal osseous defects.

AIM: The purpose of this study was to compare two types of conventional radiographic film: Ekta-speed plus and Insight (Eastman Kodak Co, Rochester NY, USA) for accuracy of measurements of approximal bone loss. METHODS AND MATERIALS: Four dried human mandibles with complete dentition were selected. Radiographic images were made with a standardized technique. Mesial and distal bone levels on the mandibular premolar and molar teeth were measured on the two types of radiographs (Ekta-speed plus and Insight) by nine observers. The data obtained by the observers were compared with the primary investigators' corresponding measurements, which served as "the ground truth" for this study. RESULTS: The results of the analysis of variance (ANOVA) demonstrated a significant difference for the Ekta-speed plus film (p = .001), but the difference was not clinically significant. The level of intra-rater reliability was high for the observers (0.969 to 0.990). CONCLUSION: Performance of Ekta-speed plus and Insight films was found to be similar.

Alexithymia and pain in temporomandibular disorder.

OBJECTIVE: To clarify the relationship of global alexithymia and its facets with pain, assessed prospectively using experience sampling methods (ESMs), in temporomandibular disorder (TMD). METHODS: People with painful TMD (n=49), pain-free somatic controls (24 people with disk displacement), and healthy controls (n = 28) completed measures of alexithymia (Toronto Alexithymia Scale-20 [TAS-20]) and depressed mood. Patients with painful TMD used ESM to record jaw pain multiple times daily for a week. RESULTS: The somatic and the healthy controls were equivalent on alexithymia and were combined. The painful TMD group had higher difficulty in identifying feelings but lower externally oriented thinking (EOT); only the latter effect remained after covarying depressed mood. Among patients with painful TMD, the TAS-20 total and EOT correlated positively with pain severity after controlling for depressed mood. CONCLUSION: Findings highlight the complex relationships of alexithymia and its facets to TMD pain. Research should examine alexithymia facets separately and distinguish between methods that compare groups on alexithymia (e.g., pain patients versus controls) and those that correlate alexithymia with pain severity within a group.

The role of parafunctions, emotions and stress in predicting facial pain.

BACKGROUND: The authors conducted a study to examine the degree to which parafunctions and emotional states predicted jaw pain in subjects with temporomandibular disorder (TMD) and control subjects. METHODS: Ninety-six subjects diagnosed with myofascial pain, myofascial pain and arthralgia, disk displacement or no TMD symptoms participated. The authors used experience sampling methodology to collect data on pain, behaviors and emotions. They paged subjects approximately every two hours, but not during sleep. When paged, subjects completed a brief questionnaire containing rating scales of jaw pain, masticatory muscle tension, time and intensity of tooth contact, mood and stress level. RESULTS: Analyses of variance showed that groups differed significantly (P < .05) in terms of pain; masticatory muscle tension; and a composite variable measuring time and intensity of contact; mood; and stress. The two myofascial pain groups scored higher on these measures than did the group with disk displacement and the control group. The authors used masticatory muscle tension, the composite variable, mood and stress to predict jaw pain using linear regression. The model was significant and accounted for 69 percent of the variance in jaw pain. Because tension was so highly correlated with jaw pain, the authors removed this variable and re-ran the analysis. The second model also was significant and accounted for 46 percent of the variance in jaw pain. CONCLUSIONS: Parafunctional behaviors, especially those that increase muscle tension, and emotional states are good predictors of jaw pain levels in patients with TMD and healthy control subjects. CLINICAL IMPLICATIONS: Treatment that helps patients reduce parafunctions, excess masticatory muscle tension, stress and emotional distress should be effective in reducing TMD pain.

Tooth contact in patients with temporomandibular disorders.

Both experimental and retrospective studies suggest a link between parafunctions and pain in temporomandibular disorder (TMD) patients. To investigate the role of parafunctions in TMD, experience sampling methodology was used as a prospective test of the hypothesis that patients with TMD have higher levels of tooth contact and tension than non-TMD controls. Three groups of TMD patients and a group of normal controls carried pagers for one week, were contacted approximately every two hours by an automated calling system, and completed questionnaires assessing tooth contact, tension, and pain at each contact. Results showed that tooth contact was much more frequent among normal controls than is commonly presumed. Patients with myofascial pain with/without arthralgia reported more frequent contact, higher intensity contact, and more tension than patients with disk displacement or normal controls. Increased masticatory muscle activity responsible for tooth contact and tension may be an important mechanism in the etiology and maintenance of the myofascial pain and arthralgia of TMD.

In vitro effect of light-cure dental adhesive on IL-6 release from LPS-stimulated and unstimulated macrophages.

The objective of this study was to measure IL-6 release from LPS-stimulated and -unstimulated macrophages exposed to extracts from fresh and aged Scotchbond Multipurpose Plus adhesive disks (5 mm in diameter by 2 mm in thickness) light cured for 10, 20, or 40 s. One set of disks was aged for 16 weeks at 4 degrees C. Extracts were prepared by incubating three disks in 1 mL of serum-free culture medium for 72 h at 37 degrees C. Then macrophages (RAW 264.7) were exposed to the extracts (6.25-50 microL) for 72 h at 37 degrees C/5% CO(2). Supernatants were analyzed for cytokine levels (ELISA), and the monolayer of cells was assessed for viability (MTT assay). Unlike adhesive disk age, curing time affected cell viability. Disk extracts cured for 10 s were more cytotoxic (p < 0.05) than were extracts from 20- or 40-s cured disks. Macrophage release of IL-6 was stimulated significantly (p < 0.01) by extracts from fresh 10-s cured disks, up to 777 pg/mL and by 2 microg/mL of LPS (1174 pg/mL). The LPS response was significantly (p < 0.05) suppressed by 50 microL of extracts, which may be related to the enhanced cytotoxicity exhibited by LPS in combination with extracts. This study has demonstrated the possibility that IL-6 release is stimulated by light-cure dental adhesive applications using 10-s curings.

In vitro mutagenicity and metabolism of the cycloaliphatic epoxy Cyracure UVR 6105.

Cyracure UVR 6105 is a cycloaliphatic epoxy monomer and has both carboxylate and epoxy groups, with the potential for rapid polymerization. It is widely used in industry for the preparation of inks, resins, coatings, and was proposed for incorporation into dental composites. The objective of this study was to determine the mutagenic potential of this chemical related to its metabolite products. Several doses of Cyracure UVR 6105 were dissolved in DMSO and subjected to the Ames Salmonella mutagenicity assay. A metabolic activation system (S9-mix) was used consisting of Arochlor-induced liver S9 homogenate enriched with NADP and glucose-6-phosphate cofactors. In contrast to studies without S9-mix, Cyracure UVR 6105 exhibited enhanced genotoxic activities with strains TA100 and TA1535 in the presence of liver S9-mix. From in vitro metabolism of Cyracure UVR 6105 with S9-mix, as used in the Ames assay, several metabolites were identified. The alcohol metabolite, 3,4-epoxycyclohexylmethanol, containing intact epoxy group was identified in the organic solvent extract. This metabolite was synthesized and proved to be mutagenic against TA100 when assayed in the presence and absence of S9-mix. Results showed that the increased mutagenicity of Cyracure UVR-6105 in the presence of liver enzymes is due to the formation of the mutagenic metabolite 3,4-epoxycyclohexylmethanol.

Comparison of panoramic radiography and panoramic digital subtraction radiography in the detection of simulated osteophytic lesions of the mandibular condyle.

OBJECTIVE: The aim of this study was to compare panoramic temporomandibular joint radiography, unenhanced and color-enhanced digital-subtraction radiography with respect to detectability of simulated osteophytic lesions of the mandibular condyle. METHODS: Three dry human skulls with no obvious temporomandibular joint pathology were selected. Four sizes of bone chips were placed on the anterior aspect of the condyle at medial, central, and lateral locations. Panoramic radiographs were made with and without the chips in place. These paired radiographs were digitized, and unenhanced and color-enhanced digital-subtraction images of the original panoramic images were obtained. Eight observers evaluated 72 randomized images of each modality for the presence or absence of simulated osteophytic lesions of the mandibular condyle, grading the images on a 100-point scale. RESULTS: A(z) values for overall diagnostic accuracy of the three imaging modalities were 0.5376 for panoramic radiography, 0.7861 for unenhanced digital subtraction radiography, and 0.7923 for color-enhanced digital-subtraction radiography. Digital-subtraction radiography improved the detection accuracy of the original panoramic films. CONCLUSION: Panoramic radiographs were significantly less accurate in the detection of simulated osteophytic lesions of the mandibular condyle than the two digital-subtraction techniques.

Inter-examiner reliability of physical assessment methods for assessing temporomandibular disorders.

The purpose of this study was to evaluate whether non-expert examiners could reliably use the physical assessment methods described in the Research Diagnostic Criteria for temporomandibular disorders (RDC/TMD). Screening examinations were performed on patients with self-reported TMD symptoms by two examiners using techniques specified in Axis I of RDC/TMD diagnostic criteria. Both examinations and diagnostic categorization were carried out independently and blindly; reliability of the examiners was tested with appropriate parametric and non-parametric techniques. Assignment of diagnostic categories based on examination data showed levels of reliability ranging from acceptable to good. Concordance for diagnostic signs was good for determining opening click, closing click, and coarse crepitus, while concordance for determining protrusive click and joint space pain categories was poor. Using the examination and diagnosis procedures specified in the RDC/TMD, conscientious nonexpert examiners can assign diagnostic labels to the major subcategories of TMD with a satisfactory level of reliability.

Effect of measurement site on the dimensional accuracy of die-forming materials and techniques.

To date, the comparisons of dimensional accuracy of various die materials has been investigated primarily by using a one-dimensional measurement on one die. However, a single measurement may not adequately represent the complex preparations that are encountered frequently in clinical dentistry. This investigation utilized a standardized model with two identical dies to measure die accuracy at four sites, using five different die-forming materials/techniques.

Statistical and clinical significance: alternative methods for understanding the importance of research findings.

Statistical significance is an important tool for interpreting a study's results, but statistical significance provides an incomplete picture of results. The likelihood of obtaining statistically significant results can be manipulated by a researcher who uses large sample sizes or who compares treatments that are expected to differ greatly in outcome. Measures of effect size provide an additional tool for understanding the results of a study and evaluating the importance of the results. Meta-analyses and estimates of clinical significance can also help clinicians properly evaluate research findings.