It is widely accepted and acknowledged that data harmonization is crucial: in its absence, the co-analysis of major tranches of high quality extant data is liable to inefficiency or error. However, despite its widespread practice, no formalized/systematic guidelines exist to ensure high quality retrospective data harmonization.To better understand real-world harmonization practices and facilitate development of formal guidelines, three interrelated initiatives were undertaken between 2006 and 2015. They included a phone survey with 34 major international research initiatives, a series of workshops with experts, and case studies applying the proposed guidelines.A wide range of projects use retrospective harmonization to support their research activities but even when appropriate approaches are used, the terminologies, procedures, technologies and methods adopted vary markedly. The generic guidelines outlined in this article delineate the essentials required and describe an interdependent step-by-step approach to harmonization: 0) define the research question, objectives and protocol; 1) assemble pre-existing knowledge and select studies; 2) define targeted variables and evaluate harmonization potential; 3) process data; 4) estimate quality of the harmonized dataset(s) generated; and 5) disseminate and preserve final harmonization products.This manuscript provides guidelines aiming to encourage rigorous and effective approaches to harmonization which are comprehensively and transparently documented and straightforward to interpret and implement. This can be seen as a key step towards implementing guiding principles analogous to those that are well recognised as being essential in securing the foundational underpinning of systematic reviews and the meta-analysis of clinical trials.
Cervical cancer screening has resulted in a decrease in the occurrence of and death from cervical cancer. The Canadian Longitudinal Study on Aging (CLSA) prospectively collected health outcomes on >50,000 individuals. We sought to identify the prevalence of Canadian female participants having never undergone cervical cancer screening and the association with social determinants of health.
Methods
We performed a cross-sectional analysis from CLSA data. The main outcome was self-report of ever having undergone a pap test. Regression analyses, controlling for the complexity of the design and covariates, evaluated the association between self-reported lifetime cervical cancer screening and social determinants of health.
Results
The population-based sample comprised 22,910 participants aged 45–85, of whom 99.8% had available information on cervical cancer screening (n=22,720). The prevalence of never having undergone a pap was 14.1%; weighted prevalence, 11.8% (95%CI 11.0–12.6). Older age(10-year) (OR 1.5, 95%CI 1.4–1.6), lower education(low vs. high) (OR 1.5, 95%CI 1.2–1.9) and low household income(low vs. high) (OR 1.7, 95%CI 1.3–2.3) were associated with absence of lifetime screening. Having a religious affiliation (OR 1.3, 95%CI 1.1–1.5) and never being married/lived in common-law (OR 1.5, 95%CI 1.2–1.9) were also associated with never having undergone screening. Notably, not having a family physician was an important contributing factor (OR 2.3, 95%CI 1.6–3.3). However, of participants who never underwent a pap test, 97% reported having a family physician.
Conclusion/Implications
Our analysis highlights inequities in access to cervical cancer screening in the Canadian context. This data can help inform targeted education and empowerment strategies to increase cancer screening uptake.
Abstract Background Several conservative (i.e., nonpharmacologic, nonsurgical) treatments exist for secondary lymphedema. The optimal treatment is unknown. We examined the effectiveness of conservative treatments for secondary lymphedema, as well as harms related to these treatments. Methods We searched MEDLINE ® , EMBASE ® , Cochrane Central Register of Controlled Trials ® , AMED, and CINAHL from 1990 to January 19, 2010. We obtained English- and non-English-language randomized controlled trials or observational studies (with comparison groups) that reported primary effectiveness data on conservative treatments for secondary lymphedema. For English-language studies, we extracted data in tabular form and summarized the tables descriptively. For non-English-language studies, we summarized the results descriptively and discussed similarities with the English-language studies. Results Thirty-six English-language and eight non-English-language studies were included in the review. Most of these studies involved upper-limb lymphedema secondary to breast cancer. Despite lymphedema's chronicity, lengths of follow-up in most studies were under 6 months. Many trial reports contained inadequate descriptions of randomization, blinding, and methods to assess harms. Most observational studies did not control for confounding. Many studies showed that active treatments reduced the size of lymphatic limbs, although extensive between-study heterogeneity in areas such as treatment comparisons and protocols, and outcome measures, prevented us from assessing whether any one treatment was superior. This heterogeneity also precluded us from statistically pooling results. Harms were rare (< 1% incidence) and mostly minor (e.g., headache, arm pain). Conclusions The literature contains no evidence to suggest the most effective treatment for secondary lymphedema. Harms are few and unlikely to cause major clinical problems.
In December 2002, a national workshop was held in Ottawa to guide research directions for the caregiving of people with Alzheimer's disease in Canada. Prior to the workshop, a search was conducted to identify Canadian-based investigators who have conducted research related to caregiving of individuals with cognitive impairment, Alzheimer's disease and other dementia, identify relevant Canadian research studies, and provide an overview of results and themes emerging from this research. This paper summarizes findings from our search for Canadian studies, research themes identified at the national workshop in Ottawa, and recommendations from the workshop. The report first outlines patterns of caregiving and subsequently focusses on the impact of different ways of assisting Alzheimer's caregivers. It concludes by outlining the methodological challenges of observational and intervention studies related to Alzheimer's caregiving and recommendations on how to increase Canada's capacity to conduct research on Alzheimer's caregiving.
We detail a new approach to the creation of normative data for neuropsychological tests. The traditional approach to normative data creation is to make demographic adjustments based on observations of correlations between single neuropsychological tests and selected demographic variables. We argue, however, that this does not describe the implications for clinical practice, such as increased likelihood of misclassification of cognitive impairment, nor does it elucidate the impact on decision-making with a neuropsychological battery.We propose base rate analyses; specifically, differential base rates of impaired scores between theoretical and actual base rates as the basis for decisions to create demographic adjustments within normative data. Differential base rates empirically describe the potential clinical implications of failing to create an appropriate normative group. We demonstrate this approach with data from a short telephone-administered neuropsychological battery given to a large, neurologically healthy sample aged 45-85 years old. We explored whether adjustments for age and medical conditions were warranted based on differential base rates of spuriously impaired scores.Theoretical base rates underestimated the frequency of impaired scores in older adults and overestimated the frequency of impaired scores in younger adults, providing an evidence base for the creation of age-corrected normative data. In contrast, the number of medical conditions (numerous cardiovascular, hormonal, and metabolic conditions) was not related to differential base rates of impaired scores. Despite a small correlation between number of medical conditions and each neuropsychological variable, normative adjustments for number of medical conditions does not appear warranted. Implications for creation of normative data are discussed.
