Info

One Minute Weight Loss

Best Weight Loss Programs That Work

Get Instant Access

r Compares programs with different outcomes

' Difficulty defining monetary value of health consequences r Provider & payer may lack understanding of tool

Fig. 5.17. Pharmacoeconomic (PE) Studies 3 & 4

diabetes treatment an antiobesity drug such as Xenical® (orlistat) and its effect on lowering the hemoglobin A1C with the cost and effectiveness of a different antiobesity drug or a structured dietary program.

Another example is provided by the biotechnology product Cerezyme® (imiglucerase), which is used for the treatment of Gaucher disease. Gaucher disease is characterized by a deficiency of beta-glucocerebrosidase activity resulting in the accumulation of glucocerebrosidase in tissue macrophages, which become engorged and are typically found in the liver, spleen, and bone marrow and occasionally in the lung, kidney, and intestine. The clinical consequences include severe anemia, thrombocytopenia, progressive hepatosplenomegaly, and skeletal complications such as osteonecrosis and osteopenia, with resultant pathological fractures. Cerezyme® improves anemia and thrombocytopenia, reduces spleen and liver size, and decreases cachexia to a degree similar to that observed with alglucerase. For many patients, enzyme replacement therapy has been effective, returning the liver, spleen, and bone marrow back to an effective degree of function. Cost-effectiveness studies of enzyme replacement therapy for Gaucher disease have consistently shown that the treatment is effective, safe, and associated with improved quality of life. On the other hand, it is expensive. Estimated cost of the enzyme alone ranges from $70,000 to $550,000 per year for a typical adult with Gaucher disease, depending on the dose. A cost-effectiveness analysis would determine if the additional cost associated with Cerezyme® treatment is matched or exceeded by its benefits compared with an alternate treatment.

A second type of pharmacoeconomic analysis is the cost-minimization analysis. It is used to define the most economical treatment among different alternatives with equal efficacy/ effectiveness and safety profiles, assumed but not directly assessed in the calculations. An example of a cost-minimization analysis is the comparison of a brand name and equivalent generic drug. A generic drug is identical, or bioequivalent, to a brand name drug in dosage form, safety, strength, route of administration, quality, performance characteristics, and intended use. Although generic drugs are chemically identical to their branded counterparts, they are typically sold at substantial discounts from the branded price. The generic drug will always show advantages by cost-minimization analysis. Two antihypertensive nongeneric products with different clinical profiles could be evaluated with this method also, but the different contribution of the side effects or administration requirements will not be incorporated even though they may be important in their use.

Five growth hormone (somatotropin) products were available in the U.S. market in 2003: Nutropin AQ® (Genetech, 5 mg $441.00); Genotropin® Injection (Pharmacia, 5.8 mg $210.00, 13.8 mg $504); Humatrope® (Lilly, 5 mg $220.50, 6 mg $264.60, 12 mg $529.20, 24 mg $ 1058.40); Saizen® (Serono, 5 mg $210.00, 8.8 mg $336.00); and Norditropin® (Novo Nordisk, 4 mg $170.40, 8 mg $352.80). Based on the cost of each drug, a cost-minimization analysis is performed to identify which of these similar products has the lowest cost while providing the same benefit as the others.

A third type of pharmacoeconomic analysis is the cost-utility analysis. This type of analysis is based on a sophisticated methodology in which benefits are calculated using parameters that take into account the quality of life of the patient. These analyses are an extension of the lifetime cost effectiveness analysis, because they estimate both quality of life and its duration. The most utilized indicator for quality of life is the quality adjusted life year (QALY), which corresponds with a year of life adjusted for its quality.

Cost-utility analysis has been used for the drug Epogen® (erythropoietin). Assume that a patient, who has renal disease and the anemia associated with it, is treated with Epogen® and has good control. That patient is assigned a utility value of 0.9 on a scale of 0 to 1. Also assume that an untreated patient with poorly controlled disease has an average utility value of 0.5. Therefore, 10 years of life of the first patient corresponds with 9 QALYs (i.e., 10 x 0.9), whereas 10 years of life for the second patient corresponds with 5 QALYs (i.e., 10 x 0.5). The QALYs are incorporated into a lifetime cost-effectiveness analysis to determine the cost utility of each therapy.

The fourth type of pharmacoeconomic analysis is the cost-benefit analysis. When both costs and benefits of a treatment are measured in monetary values, cost-benefit analysis is a useful tool. Future costs and benefits are discounted to their current value and take into account the "time value of money." Because of inflation, a dollar today is not equivalent to a dollar in the future. However, the application of cost-benefit analysis in pharmacoeconomics is limited, due to the difficulties in assigning a monetary value to health outcomes and a patient's life. For example, when evaluating the cost of "statin" drugs versus the consequences of not treating patients with hypercholesterolemia, the cost associated with developing cardiac disease, a stroke, or death must be given a value benefit in dollars. Further, the statins may have additional benefits unrelated to reducing cholesterol levels, which are not measured.

Another example is provided by the drug Enbrel® (etaner-cept). The cost of a new biotechnology agent, such as Enbrel® for the treatment of rheumatoid arthritis, may be higher than other available agents due to its innovative mechanism of action. Not only does Enbrel® stop and relieve the pain associated with this form of arthritis, but unlike most other drugs used in this setting, such as nonsteroidal anti-inflammatory agents, it also stops joint erosion, improves mobility, and improves quality of life. As exemplified by Enbrel®, new therapies developed through biotechnology can be of great value as long as the benefits exceed the costs. Consequently, it is critically important for new biotechnology products to identify and quantify all the benefits they offer over current treatment options. Benefits may include improved outcome or efficacy including stopping and/or reversing disease progression, reduced side-effects or complications, reduced hospitalizations or bed-days, improved quality of life, improved morbidity and mortality, and reduced total health care costs. Well-designed pharmacoeconomic analyses can be instrumental in defining the overall benefits of these new therapies [8-12].

Quality of life was briefly discussed in the section on cost-utility analyses. In contrast with efficacy, safety, and cost-effectiveness studies, which are viewed by providers, investigators, and researchers as important in the decision-making process associated with drug development, quality of life (QOL) studies are often viewed as supplemental (Fig. 5.18). An exception is the role in the development of biologicals, most likely due to their higher costs [8, 11,18].

r Efficacy, safety, and cost-effectiveness studies are all viewed as important in decision making r In contrast, QOL studies are often viewed as supplemental (However, they are key parameters with biologicals)

r QOL studies-disease specific (FACT in cancer) vs. generic (SF-36 health survey)

r What role does QOL play in drug trials and formulary decisions?

r Who benefits from QOL data?

r Who should pay for better QOL information?

Fig. 5.18. Quality of Life Studies

Although many studies provide good QOL data demonstrating additional significant benefits for patients, and their impact on the decision-making process has not been well studied. Some findings are that the role of QOL information in influencing managed care decision-making is not well understood, because research on the subject is relatively new and/or has been minimal, designs are less well understood and accepted, gold standards are not as well recognized, and applicability to specific health care settings may be missing [8, 11, 18].

For QOL studies, one can define health as "not merely the absence of disease, but complete physical, psychological and social well-being." To measure QOL, multiple tools have been developed and validated. There are generic instruments such as the SF-36 [13] and disease-specific instruments, such as St. George's Respiratory Questionnaire for COPD (chronic obstructive pulmonary disease) [14] and the Functional Assessment of Cancer Therapy (FACT) questionnaire for cancer [15]. Such generic measures have questionable applicability to certain diseases or to a drug's impact on the disease or sensitivity to pick up specific disease changes in QOL, leading to a need to develop such disease-specific instruments. However, the disease-specific tools must be repeatedly used and validated before acceptance by the medical community and health care systems.

Generic QOL instruments are used for a wide range of diseases to determine how treatment influences day-to-day activities, well-being, and social functioning. Generic instruments can be used to compare the impacts of different diseases. The SF-36 is a health survey with 36 items constructed to identify a patient's health status. It was designed for use in clinical practice and research, health policy evaluations, and general population surveys. The SF-36 includes one multi-item scale that assesses eight health concepts: (1) limitations in physical activities because of health problems; (2) limitations in social activities because of physical or emotional problems; (3) limitations in usual role activities because of physical health problems; (4) bodily pain; (5) general mental health (psychological distress and well-being); (6) limitations in usual role activities because of emotional problems; (7) vitality (energy and fatigue); and (8) general health perceptions. The survey was constructed for self-administration by persons 14 years of age and older or for administration by a trained interviewer in person or by telephone.

Disease-specific QOL instruments are usually more responsive to changes in QOL than generic and utility measures. Disease-specific QOL tools are more specific for disease but less applicable for drug formulary decision-making. Disease-specific QOL instruments require validation and applicability to the disease and disease treatment in routine clinical practice. They also must include practical measures that can generate reproducible results.

The Asthma Quality of Life Questionnaire (AQLQ) is a 32-item questionnaire that has been developed to measure the functional impairments that are most important for adults

(17-70 years) with asthma [16]. A pediatric version is also available [17]. The items are in four domains (symptoms, emotions, exposure to environmental stimuli, and activity limitation). The instrument is in both interviewer- and self-administered formats and takes approximately 10 minutes to complete at the first visit and 5 minutes at follow-up. Several independent studies have demonstrated the strong evaluative and discriminative measurement properties and validity of the Asthma Quality of Life Questionnaire. It has been used successfully in a large number of clinical trials and in clinical practice around the world.

Another example of a disease-specific QOL is the Functional Assessment of Cancer Therapy (FACT) [15]. An assessment of fatigue may consider broader concerns, such as global quality of life and symptom distress. Some of the fatigue scales, such as the unidimensional three-item scale of the EORTC QLQ-C30 and the multidimensional fatigue sub-scale of the Functional Assessment of Cancer Therapy (FACT), are themselves modules of well-validated quality of life instruments. The larger scale may be included if additional evaluation of quality of life is valuable. For the other fatigue scales, a separate quality of life questionnaire will be needed to accomplish the same goal. Most patients with cancer or AIDS have multiple symptoms. Fatigue, pain, and psychological distress are the most prevalent in most populations. Given the likelihood of multiple symptoms, it may be informative to add a measure of symptom prevalence and distress to the fatigue-assessment strategy. This approach also can clarify the extent to which fatigue associates with other symptoms.

Although many studies provide good quality of life data demonstrating additional significant benefits for patients, the impact of the data on drug approval and on formulary decision-making is uncertain.

Who benefits from QOL studies? Pharmacoeconomics and QOL information is increasingly discussed now in formulary and drug use decisions. However, researchers have been unable to identify the extent of influence that pharmacoeco-nomics and QOL information has on formulary decision making. Pharmacy and medical directors in health care systems historically focused on standard clinical parameters of safety and efficacy or cost (cost-effectiveness or cost of treatment) in their decision-making process. This is largely due in part to the nature of managed care's focus on reducing cost. The concept of health care insurance or coverage was based on providing services for medical necessity. How does QOL fit into the puzzle of medical need? Should health care be responsible for providing care, services, or products that will improve the overall well-being of patient?

Who should pay for QOL? Patients reap the benefits of services or products that improve their QOL. If patients are reaping almost all the benefits, then should they be accountable to pay for these services or products? Consumers will readily pay for items that provide convenience or improve their quality of life (i.e., dishwashers and washing machines, housekeepers or gardeners). Some consumers are willing to r Understanding impact of a product on health care system and incorporating that information into study design:

p Clinical efficacy and safety vs. effectiveness c Improved outcomes (morbidity and mortality) and QOL r Health care utilization (efficiency) vs. efficacy and safety:

p Study Differences: setting, design, population, patiententry, intervention, outcomes, generalizability, confounding factors r Separate from clinical trials in NDA: c Potential negative impacton NDA/BLA o Many healthcare settings and perspectives p Inaccessibility of data and high cost of data acquisition p Not needed for approval by regulatory authority p Needed for health system use of product

Fig. 5.19. PE Drug Development Challenges pay $6.00 a pill to improve their quality of life but will not pay $2.00 a pill to prevent them from dying of a heart attack. Are there ways or means to quantify and translate these benefits into the health care system? Other QOL studies, as in anemia in renal disease and Epogen® (epoietin alpha), used QOL as the primary end points in product approval, and patients had (have) substantial and exceptionally dramatic benefits in daily living activities such that Medicare decided to pay for the product. In order to make QOL more valuable in the decision-making process, future studies need to define more fully the economic value of QOL in the health care system [8-12].

In addition to the benefits they may provide during the drug development process in terms of added significant study end points and patient care benefits, pharmacoeconomic studies also offer a number of challenges in their conduct (Fig. 5.19). For example, it is essential to consider the potential impact of a product on the entire health care system and incorporate this additional information into the study design. It is also important to separate, when possible, pharmacoeconomic studies from those clinical trials required as part of the NDA/BLA submission because they may have a negative impact on the submission. Phase 3 studies do not use the typical patients that you find in health care systems, such as managed care organizations (MCO), and you want to use in PE studies. Furthermore, study design differences for PE versus clinical r Lack of understanding of applications of PE studies r Sub-optimal use of health outcome and PE data by health care systems r Studies or analyses needed: p Easier to understand p Relevant to health systems r Clinical studies lack PE content r Unique settings impacting PE assessments r Lack of all necessary data available r Lack of integrated health care data

Fig. 5.20. PE Application Challenges

Example: Cancer Patient (ALL) r Drug Issue - Liver Metabolism: (+ or - activity)

p e.g., CYP2D6, TPMT r Drug Issue - Patient's Receptor Sensitivity: (> or < action)

p e.g., P1AR, PXR, GR, VDR r Cancer Genotypes: (> or < disease; > or < response) p Disease subtypes, e.g., Her2Neu, Bcr/ Abl, P53, BEX, VEGF p Drug resistance factor, e.g., MDR r Patient's Infection Defense:

p Immune system, e.g., IL1, IL6, TNF, IL2, MHC/ HLA r Results: p > or < Response? and/ or Toxicity? p Outcomes - composite of all the pharmacogenetic changes

Fig. 5.21. Gene Variability in Pharmacogenetics studies is quite different (e.g., setting [MCO vs. university hospital], patient entry [all comers in a health system for PE, inclusive vs. exclusive], intervention [specific drug at specific doses vs. standard of care at these institutions], and outcomes). Health care settings best used for phase 3 studies may not have the type of patients or data needed for PE trials. Training of investigators and patient monitors is a huge challenge in time, costs, and reliability of the QOL information. As noted earlier, clinical trials in a development plan are needed for approval and PE or QOL are not [8-12].

Although pharmacoeconomic (PE) and quality of life (QOL) information is being increasingly discussed in formulary and drug use decisions, it has been difficult to identify the extent to which this information influences formulary decision making (Fig. 5.20). Challenges to drug development regarding PE and QOL studies include which ones are required, the designs, their conduct, and their relationship to clinical studies. The industry must challenge itself to perform those studies that are as relevant as possible to the appropriate health settings, use easy to understand methodologies, and publish the information that is most important to health care providers, health systems, and payors. Also, gold standards in study design and application of the data do not generally exist for PE studies, especially given the many different types of studies and varied settings for drug use. When PE and QOL studies are done, the company needs to assist providers and payors in these settings to understand how this information, which may originate from a different setting, fits their institutions and systems. Because the clinical trials usually lack the PE or QOL data, the applicability and integration of both the clinical studies and the PE or QOL studies for a new product need to be addressed to also assist the payors and providers [9-12, 18].

Pharmacogenetics is a relatively new and complex discipline based on heritable or acquired genetic differences between groups of people that can change a drug's actions in the body (Fig. 5.21). About 60,000 single nucleotide polymorphisms exist on the coding regions of the human genome, and about 1.5 million exist in the full genome, creating a plethora of

c Identify therapy that will have high likelihood of success in groups of patients o Achieve improved individual responses o Reduce use of ineffective treatments c Reduce adverse events c Reduce cost of drug development with more efficient trials "Individualized Therapy"

Fig. 5.22. Pharmacogenetic Studies potential differences between patients' biology. This figure suggests the scope and some of the complexity of genetic variations in a cancer patient with acute lymphocytic leukemia (ALL). The cancer genotype, especially related to surface antigens, will vary in patients with the same disease and alter patient response to therapy, which is now well documented for aggressive breast cancers and Herceptin® (trastuzumab), acute myelogenous leukemia and Mylotarg® (gemtuzumab), and colorectal cancer and Erbitux® (cetuximab). Host susceptibility has genetic variation, as well as infection defense mechanisms. Drug metabolism is particularly effected by genetic variation in liver enzymes and drug clearance. In the pediatric cancer, ALL, the appropriate use (dose) of thiopurine is very dramatically changed downward tenfold by genetic variation, potentially leading to possibly fatal toxicities [19-22].

Increasing emphasis is being placed on "personalized medicine." The major goals of pharmacogenetic studies in drug development are to identify therapies that will have a high likelihood of success in individual patients and/or reduced toxicity (Fig. 5.22). An improved drug responsiveness has been demonstrated in a subpopulation of breast cancer patients with particularly aggressive cancer; that is, Herceptin® therapy significantly increases cure rates in patients with her2neu oncogene in about 25% of breast cancer patients. Another goal is to reduce use of treatments that would be ineffective in a subgroup of patients that we would know would not respond to the treatment. The current alternative is using a drug in a 100 patients, in which the response rate is 50%, but we do not know which 50 patients will be responders. Better dose selection would be possible with either less toxicity or better efficacy through genetics (e.g., thiopurines in cancer and narcotic analgesics, respectively). Reduction in the cost of drug development could be an outcome with more efficient trials; that is, products are only used in smaller groups of patients with higher likelihood of response rates to even higher degrees [19-22].

There are several potential disadvantages in incorporating pharmacogentics studies into drug development process (Fig. 5.23) [18-21]. The diagnostic use of genetics is not yet commonplace, related to, for example, the lack of knowledge of impact of genetics in many diseases, cost of tests, availability and reliability of tests, and unknown reimbursement by payors.

r Potential disadvantages: o Smaller target population with reduced sales o Cost of genotyping o Additional patient consent r Unclear clinical significance o Diagnostic and assay dilemma o Ethics including impact on insurance coverage

Fig. 5.23. Pharmacogenetic studies r Term "Compassionate" is not in IND regulations. Emergency Use and Treatment INDs

' Emergency Use - use of investigational drug or biological product in life-threatening situation when no standard acceptable treatment is available, and there is insufficient time to obtain IRB approval.

r Allows for one emergency use with out prospective IRB review. Any subsequent use of investigational product at institution requires prospective IRB review and approval.

r Company provides product after phase 2 or likely phase 3 with regulatory consent for this procedure to be done

Fig. 5.24. Compassionate Use

Other disadvantages include a smaller target population for only the genetically likely responders with reduced sales, the cost of genotyping, a need for additional and frequently separate patient consent, unclear clinical significance of phar-macogenetics to disease pathogenesis and product pharmacology, and possible ethical issues including an impact on insurability. Furthermore, large epidemiologic studies examining the associations of pharmacogenetics to diseases and with drugs are needed, which is a huge expense. This deficit is starting to be addressed in NIH funding.

The FDA has published a voluntary guidance for companies regarding the use of pharmacogenetic studies in the drug development process, their role in the approval process, and how to submit the data for its review [23].

The term "compassionate" is not in the IND regulations. "Compassionate use" studies are either "emergency use" protocols or "Treatment INDs" (Fig. 5.24). The emergency use provision governs the use of an investigational drug or biological product in a life-threatening situation when no standard acceptable treatment is available and in which there is insufficient time to obtain institutional review board (IRB) approval before treatment must be started. This provision allows for one emergency use without prospective IRB review. The use must be reported to the IRB according to federal and local requirements. Any additional use of the investi-gational product requires prospective IRB review and approval. This emergency use is normally only done after phase 2 is r Make new drugs available to desperately ill patients early in drug development process.

r Need preliminary evidence of drug efficacy & safety, documentation drug is intended to treat a serious or life-threatening disease, and no alternative therapy available to treat that stage of disease in intended patient population.

r Patient is not eligible to be in definitive clinical trials, which usually must be well underway (e.g., during phase 3), if not almost finished.

r Enables FDA to obtain additional data on drug's safety and effectiveness.

Fig. 5.25. Treatment IND

" Definition:

O Observational studies of larger size

O Prospective or retrospective

O Longitudinal over specified time

O Hard end point, well-defined

O Sites more in community settings

O Comparator control groups

O Information sources: databases, patients, registries, charts/medical histories

O More representative population in real world r Goals:

O Study target disease e.g., risk groups, disease descriptors, practice patterns, market/population size

O Estimate rates of background events, e.g., adverse events O Design large simple post-marketing safety trials

Fig. 5.26. Epidemiology Studies - Definition and Goals complete and ideally phase 3 is done or almost complete, so that a reasonable idea of both safety and efficacy exist.

For a company with such a life-saving product, emergency use is not often able to be accomplished, because of the time required to document the patient's need and diagnosis, investigator's/ practitioner's credentials, and the distribution requirements for the product. These issues are not vicarious requirements from a company but minimum regulatory and especially safety issues. Usually when it is done, a protocol is created in advance to cover this usage including approval by regulatory authorities. The necessary inclusion and exclusion criteria created for this protocol can present a barrier to such open-ended use, because the individual patient and family may have an expectation of availability of the product, but the patient may not qualify. This situation can become a possible public relations boon or fiasco, which is a practical challenge to control expectations.

The Treatment IND provision makes new drugs available to desperately ill patients early in the drug development process (Fig. 5.25). Approval of a treatment IND requires preliminary evidence of drug efficacy, documentation the drug is intended to treat a serious or life-threatening disease, there is no alternative therapy available to treat that stage of the disease in the intended patient population, and the patient or patient population is not eligible to be in the definitive clinical trials. The clinical trials program usually must be well underway, (e.g., during phase 3, if not almost finished). A Treatment IND also enables the FDA to obtain additional data on safety and effectiveness. A protocol must be written by the company and approved by the regulatory authorities for this usage. Also, regulatory provisions allow a company to charge the health care system for this usage, but the company needs to share costs of production and research costs with the regulatory authority, which is proprietary information.

Epidemiology studies use observational study designs in large populations (e.g., hundreds to thousands of patients) to improve our understanding of diseases and therapies (Fig. 5.26). The sources of information about the patients, diseases, treatments, and events include large databases, such as Medicaid claims data, patient interviews (in person, telephone, mail, or Internet), patient registries, and medical record reviews. Each data source has its limitations, which will qualify the results and conclusions. For example, databases can be influenced by restrictions in formulary status, treatment guidelines in place, or age of the population exposed. Interviews are susceptible to patient memory lapses and their reliability as historians. The end points are definitive (e.g., hospitalization, death, heart failure, or gastrointestinal bleed). The data can be gathered retrospectively (e.g., chart reviews) or prospectively. The patients for epidemiology studies are found most often in community practice settings, and as a result they are more representative of the "real world." Control groups, used for comparison, are usually drawn from the same population as the patients exposed to the disease and or treatment. These characteristics should result in a representative population sample, studying typical patients receiving typical treatments in typical health care settings.

The goals of these observational studies are threefold: (1) to study the target disease, which can provide information on patients at risk for an exposure or reaction, disease or patient descriptions to be used for inclusion or exclusion criteria in other studies, practice patterns in diagnosis and therapy, and population (market) sizes; (2) to estimate rates of background events, especially adverse events, helping to identify reactions in a population and the influence of disease, risk factors, or treatment; and (3) to design large simple postmarketing surveillance trials for safety assessments.

Two study designs are predominant in epidemiology research: cohort and case-control (Fig. 5.27). A cohort is a group of patients with similar characteristics, also described as patients with an exposure to a specific product in drug studies. Two cohorts, with and without exposure (often a drug), are followed over a specific time period and compared for adverse events or practice patterns or to estimate a specific reaction, which may be rare and difficult to quantify in smaller randomized trials. Case control design involves a group of cases as defined by an exposure and set of characteristics and r Cohort design:

^ Group of patients with exposure to specific products r Case control design: c Cases of patients with a specific exposure or reaction and matched controls from same population r Biases to prevent; limits to observational studies: p Sampling vs Selection vs Measurement vs Confounding p Information: subject not remember information p Reverse causality: exposure related to outcome p Detection: preferential diagnosis or selection of exposed subject o Healthy patient: health status influences outcomes o Channeling: disease severity masks drug-disease association p Confounding: Variable must be associated with exposure and outcome, and can not be an effect of exposure

Fig. 5.27. Epidemiology Studies - Design Issues r 10,000-30,000 substances identified in basic research r 100-200 reach chemical synthesis and screening r 5-10 undergo pre-clinical testing r 2-5 enter clinical trials r 1 is approved and marketed

Fig. 5.28. Failure Rates of New Chemical Entities k 33% enter phase 2

k 27% enter phase 3

r 20% undergo FDA review r Not all that undergo review are approved

Fig. 5.29. Failure Rates of INDs a group of matched control subjects, both selected from the same population.

Observational studies have potential biases that must be considered and either dealt with in the design and/or used as qualifications to the results and conclusions. Information bias involves missing information because patients do not remember events or data are missing from charts. Reverse causality bias is when the exposure (drug) is unknowingly used to treat an adverse event related to the outcome, such that epidemiologists will define exposures where timing does not coincide with the outcome. Detection bias occurs when an outcome is preferentially diagnosed in subjects who are exposed to a drug associated with the outcome. Matching well the patient characteristics, diagnoses, and other nonstudy exposures will help minimize this problem. Healthy patient bias is seen when a patient's health status (e.g., exercise or diet) influences the outcome and biases the result. One compensates for this problem through study design and observation or statistical analysis with stratification. Channeling bias occurs when the severity of a disease either masks or enhances the association between a drug exposure and the disease. Mitigation of this bias requires knowledge of the disease and modification of the study design. Confounding bias occurs when an external variable is mixed with the exposure and influences the outcome under study. Epidemiologists deal with such potential bias by stratification by the confounding variable or statistical analysis using multivariate analysis.

Was this article helpful?

0 0
Weight Loss All Star

Weight Loss All Star

Are you looking to lose weight or even just tone up? What is stopping you from having the body you want and help you feel great at the same time? I created Weight Loss All-Star for all those wanting to lose weight, and keep the weight off. I know how hard it is to do diets and stick with them, and get the motivation to get up and exercise.

Get My Free Ebook


Post a comment