Complex sampling: implications for data analysis

Investigators in dental public health often use strategies other than simple random sampling to identify potential subjects; however, their statistical analyses do not always take into account the complex sampling mechanism. Often it is not clear whether a given strategy requires adjustment for stratification and/or cluster sampling of observations. We propose that the need for such adjustment depends on the primary study objective. As a general rule, we recommend that if the study goal is to estimate the magnitude of either a population value of interest (e.g., prevalence), or an established exposure-outcome association, adjustment of variances to reflect complex sampling is essential because obtaining appropriate variance estimates is a priority. However, if the study goal is to establish the presence of an association, especially in a preliminary investigation of novel conditions or understudied populations, obtaining appropriate variance estimates may not be of primary importance; hence, adjustment of variances for complex sampling is not always required, but often is recommended. This paper describes several types of complex sampling designs, methods of adjusting for complex sampling strategies, examples illustrating the effect of adjustment, and alternative approaches for analysis of complex samples.     

Proximal caries detection: Sirona Sidexis versus Kodak Ektaspeed Plus

This study compared the accuracy of intraoral film and a charge-coupled device (CCD) receptor for proximal caries detection. Four observers evaluated images of the proximal surfaces of 40 extracted posterior teeth. The presence or absence of caries was scored using a five-point confidence scale. The actual status of each surface was determined from ground section histology. Responses were evaluated by means of receiver operating characteristic (ROC) analysis. Areas under ROC curves (Az) were assessed through a paired t-test. The performance of the CCD-based intraoral sensor was not different statistically from Ektaspeed Plus film in detecting proximal caries.     

The effect of educational intervention on the adoption of nickel-titanium rotary instrumentation in a Public Dental Service

AIM: To study the influence of two educational programmes on the adoption of nickel-titanium rotary instrumentation (NTRI) amongst general dental practitioners in a short-term as well as a long-term perspective. METHODOLOGY: All dentists employed in the Gothenburg Dental service (n = 148) were enrolled in the study. The clinics in the organization were randomly assigned to one of two educational programmes. In the first programme a 4-h lecture on root canal instrumentation was given. In the second programme the lecture course was supplemented by a 6-h hands-on training session. The short-term effect was measured by a questionnaire distributed 6 months after completed education. The long-term effect was evaluated 4 years later. RESULTS: The overall utilization rate of NTRI increased from 4% to 73%. However, lectures in combination with hands-on training resulted in a better short-term acceptance rate (94%) than if teaching was given only in lecture-format (53%) (P = 0.000). As a consequence, all staff were offered hands-on training. The long-term adoption rate was 88%. Reasons for accepting the new technology usually were found within the 'relative advantage' category. Common reasons for dentists not to adopt NTRI were that they could not get started or that they found no advantage over the old technology. CONCLUSIONS: The short-term adoption of a new technology might be influenced by the design of an introductory educational programme. For clinical procedures, such as root canal instrumentation, the inclusion of hands-on training sessions seems to be important to reach a high acceptance rate.     

Relationship between number of proximal contacts and survival of root canal treated teeth

AIM: The present study tested the hypothesis that having two proximal contacts (PCs) at access is associated with improved survival of root canal treated (RCT) teeth, controlling for important presenting conditions, endodontic variables and restorative factors. METHODOLOGY: A treatment database at the University of Iowa College of Dentistry was used to identify permanent teeth undergoing obturation between 1 July 1985 and 31 December 1987. The list was restricted to teeth of patients with at least one dental visit in each two-year interval from 1985 to 1996, and a simple random sample of 280 patients (n = 400 teeth) was selected. Dental charts, radiographs, and computerized databases were examined to ascertain variables of interest and to verify study inclusion criteria. Kaplan-Meier survival estimates were generated for the 221 teeth satisfying study inclusion criteria. Multivariate Cox models were developed, with standard errors adjusted to account for clustering of teeth within patients. RESULTS: The final Cox model showed that teeth with < or =1 PC at access were lost at a rate three times that of teeth with 2 PCs (hazard ratio = 3.1; 95% confidence interval = 1.9-5.1), controlling for tooth type, presence of radiographic caries at access, and presence of a crown before or after obturation. CONCLUSIONS: Because RCT teeth with two PCs at access experienced substantially better survival than teeth with fewer than two PCs, the influence of PCs on prognosis should be recognized during treatment planning. Future research should employ prospective study designs, capture additional variables, and provide data to support endodontic treatment decisions.     

Teeth with large amalgam restorations and crowns: factors affecting the receipt of subsequent treatment after 10 years

BACKGROUND: The authors evaluated the factors associated with the receipt of subsequent treatment by teeth restored with a large amalgam restoration or a large amalgam restoration and crown restoration after 10 years. METHODS: The authors used retrospective data from the University of Iowa College of Dentistry (Iowa City, Iowa) administrative database and patient records to evaluate patient and tooth factors for their association with the two primary outcomes: receipt of any subsequent treatment and receipt of catastrophic treatment (extraction, endodontic therapy). RESULTS: The authors followed 518 teeth over a 10-year period (49 percent with large amalgam restorations and 51 percent with crowns). Sixty-four percent of the large amalgam restorations and 32 percent of the crowns received subsequent treatment during the 10 years. In addition to restoration type, the patient's sex, history of grinding teeth and having a broken tooth were related to the tooth's receiving subsequent treatment. Twenty-two percent of large amalgam restorations and 12 percent of crowns received catastrophic treatment with the odds of teeth with large amalgam restorations receiving a catastrophic treatment being 2.1 times the odds of teeth with crowns receiving catastrophic treatment. CONCLUSIONS: Teeth with crowns were less likely to receive any treatment or catastrophic treatment over 10 years than were teeth with large amalgam restorations. Patient and tooth factors also were related to a tooth experiencing subsequent treatment. CLINICAL IMPLICATIONS: Teeth with crowns received less subsequent treatment than teeth with large amalgam restorations. This could be related to both the difference in longevity between the two restorations, as well as how appropriately treatment was planned for each procedure. Cost differences between the two restorations need to be factored into the decision-making process.     

Composite score for prediction of 30-day orthopedic surgery outcomes

The LACE+ (Length of stay, Acuity of admission, Charlson Comorbidity Index score, and Emergency department visits in the past 6 months) risk-prediction tool has never been tested in an orthopedic surgery population. LACE+ may help physicians more effectively identify and support high-risk orthopedics patients after hospital discharge. LACE+ scores were retrospectively calculated for all consecutive orthopedic surgery patients (n = 18 893) at a multi-center health system over 3 years (2016-2018). Coarsened exact matching was employed to create “matched” study groups with different LACE+ score quartiles (Q1, Q2, Q3, Q4). Outcomes were compared between quartiles. In all, 1444 patients were matched between Q1 and Q4 (n = 2888); 2079 patients between Q2 and Q4 (n = 4158); 3032 patients between Q3 and Q4 (n = 6064). Higher LACE+ scores significantly predicted 30D readmission risk for Q4 vs Q1 and Q4 vs Q3 (P < .001). Larger LACE+ scores also significantly predicted 30D risk of ED visits for Q4 vs Q1, Q4 vs Q2, and Q4 vs Q3 (P < .001). Increased LACE+ score also significantly predicted 30D risk of reoperation for Q4 vs Q1 (P = .018), Q4 vs Q2 (P < .001), and Q4 vs Q3 (P < .001).     

The relationship between quadriceps angle and tibial tuberosity–trochlear groove distance in patients with patellar instability

Purpose

The quadriceps angle (Q-angle) represents the angle between the vector of action of the quadriceps and the patellar tendon. An increased Q-angle has been associated with an increased risk of patellar instability, although there is disagreement on its reliability and validity as it is affected by the position of the limb and contraction of the quadriceps. Tibial tuberosity?Ctrochlear groove distance (TT?CTG) is ascertained by axial CT scanning, with an increased value associated with patellar instability. This study aimed to determine whether the Q-angle correlates with the TT?CTG distance in patients with patellar instability.

Methods

Q-angles were measured in 34 knees that had previously undergone CT scanning for assessment of patellar instability. Measurements were made with the patient supine, the knee extended and the lower limbs in neutral rotation with the quadriceps relaxed and contracted. TT?CTG distance was measured on CT scanning in an identical position.

Results

Of the 34 knees measured, 24 had symptoms of patellar instability, and 10 were normal. A significant negative correlation between relaxed Q-angle and TT?CTG in all knees was demonstrated (p?=?0.028). In symptomatic knees, contracted Q-angle also demonstrated a significant negative correlation with TT?CTG (p?=?0.037).

Conclusions

If TT?CTG distance is regarded as the gold standard measurement, Q-angle is not a reliable indicator of patellar instability. There is a clear need to develop methods to more fully characterise the knee and factors contributing to patellar instability.

Level of evidence

II.