Observational studies and randomized trials

Personalizing Behavioral Interventions Through Single-Patient (N-of-1) Trials

Behavioral interventions are typically studied with the use of a conventional between-subject randomized controlled trial (RCT) design. In this design, the effect of an intervention on one group of patients is compared with the effect of a control condition on another group of patients, such that a between-subject change is tested. A between-subject design has an underlying assumption that there is a homogenous treatment effect for a behavioral intervention, drug or psychotherapy, and that the way the intervention operates in the study that will tend to operate in the same way in many other patients. We review some of the philosophical and practical problems with the use of this design when a clinician is attempting to decide on a course of behavioral treatment aimed at within-subject change in patients who are likely to have heterogeneous or unique responses to behavioral treatment. We also review the biases inherent in our current clinical practice model, which does not use any empirical data collection or design for testing if a treatment is useful, and also in the conventional between-subject personalized medicine RCT designs. We propose increased use of single-patient (also known as N-of-1) trials that employ within-subject designs, in cases where treatment response is heterogeneous-as is the case for most psychological and behavioral treatments. Limitations of such designs include that they can only be used when the treatment is potentially reversible, the patient can act as their own control, and the outcome can be measured repeatedly. Increased use of within-subject trials may address in many more instances the more clinically relevant question of how a specific patient will respond to a specific treatment, and could introduce a more harmonious scientific approach into the way we treat our patients. We have incorporated a case presentation that illustrates the complexities of applying evidence drawn from these different designs to selecting and evaluating treatments for the behavioral issues commonly faced by clinicians and patients.

Retrospective database analysis on the effectiveness of typical and atypical antipsychotic drugs in an outpatient clinic setting

OBJECTIVE

To report the outcomes of a retrospective database analysis to compare the effectiveness of atypical and typical antipsychotic drugs.

METHODS

Medical records of patients admitted to the psychiatry outpatient clinic between January 1998 and October 2005 were retrospectively reviewed. Data obtained from patient records were noted on a special form assessing four aspects of the treatment history: socio-demographic features, disease characteristics, initial treatment at the time of admission, and course of treatment. Patient groups (typical/atypical and Risperidone/Haloperidol/Olanzapine) were compared for time to all-cause medication discontinuation and rate of discontinuation.

RESULTS

There was no statistically significant difference in the duration of treatment between patients using atypical (n = 150) and typical (n = 124) antipsychotics. The duration of treatment was significantly longer in patients on Haloperidol (n = 91) compared with those on Risperidone (n = 63). Rates of discontinuation over 18 months were 59.3% for patients on atypical antipsychotics and 57.3% for those on typical antipsychotics, and 68.3% for patients on Risperidone, 51.6% for patients on Haloperidol and 54.3% for patients on Olanzapine.

CONCLUSION

Despite our hypothesis patients with chronic schizophrenia discontinued their atypical and typical antipsychotics, at a high rate with no significant difference indicating substantial limitations in the effectiveness of these drugs.

Is it more effective for highly trained swimmers to live and train at 1200 m than at 1850 m in terms of performance and haematological benefits?

OBJECTIVES

The effects of living and training have not been compared at different altitudes in well trained subjects.

METHODS

Nine international swimmers lived and trained for 13 days similarly at 1200 m (T1200) and 1850 m (T1850). The two altitude training periods were separated by six weeks of sea level training. Before and after each training trip, subjects performed, at an altitude of 1200 m, an incremental exercise test to exhaustion of 5 x 200 m swims and a maximal test over 2000 m.

RESULTS

There was no difference in Vo(2)max after each training trip: the before values were 58.5 (5.6) and 60.4 (6.7) ml/kg/min and the after values were 56.2 (5.2) and 57.1 (4.7) ml/kg/min for T1200 and T1850 respectively. The 2000 m performance had improved during T1200 (1476 (34) to 1448 (45) seconds) but not during T1850 (1458 (35) v 1450 (33) seconds). Mean cell volume increased during T1850 (86.6 (2.8) to 88.7 (2.9) microm(3)) but did not change during T1200 (85.6 (2.9) v 85.7 (2.9) microm(3)). The proportion of reticulocytes decreased during T1200 (15.2 (3.8)% to 10.3 (3.4)%) and increased during T1850 (9.3 (1.6)% to 11.9 (3.5)%).

CONCLUSIONS

The short term effects of 13 days of training at 1200 m on swimming performance appear to be greater than the same type of training for the same length of time at 1850 m. As mean cell volume and proportion of reticulocytes only increased during training at 1850 m, the benefits of training at this altitude may be delayed and appear later on.

Methodological aspects in the assessment of treatment effects in observational health outcomes studies

Prospective observational studies, which provide information on the effectiveness of interventions in natural settings, may complement results from randomised clinical trials in the evaluation of health technologies. However, observational studies are subject to a number of potential methodological weaknesses, mainly selection and observer bias. This paper reviews and applies various methods to control for selection bias in the estimation of treatment effects and proposes novel ways to assess the presence of observer bias. We also address the issues of estimation and inference in a multilevel setting. We describe and compare the use of regression methods, propensity score matching, fixed-effects models incorporating investigator characteristics, and a multilevel, hierarchical model using Bayesian estimation techniques in the control of selection bias. We also propose to assess the existence of observer bias in observational studies by comparing patient- and investigator-reported outcomes. To illustrate these methods, we have used data from the SOHO (Schizophrenia Outpatient Health Outcomes) study, a large, prospective, observational study of health outcomes associated with the treatment of schizophrenia. The methods used to adjust for differences between treatment groups that could cause selection bias yielded comparable results, reinforcing the validity of the findings. Also, the assessment of observer bias did not show that it existed in the SOHO study. Observational studies, when properly conducted and when using adequate statistical methods, can provide valid information on the evaluation of health technologies.

Invitation to a dialogue between researchers and clinicians about evidence-based behavioral medicine

BACKGROUND

Evidence-based behavioral medicine (EBBM) aims to improve the process through which best scientific research evidence can be obtained and translated into best clinical decisions regarding behavioral treatments to improve health.

PURPOSE

The objective was to examine some legitimate concerns raised by both clinicians and researchers about the evidence-based movement.

METHODS

This article begins with a discussion of clinicians' fears that EBBM devalues clinical judgment and the therapist-patient relationship, will be used to restrict practice, is unnecessary, and is based on research that is irrelevant to clinical decision making. Next we consider researchers' worries that EBBM neglects evidence not based on randomized controlled trials and ignores causal mechanisms.

RESULTS

We find that these fears, although understandable, largely reflect misinterpretations of the evidence-based movement. Further, it is suggested that behavioral medicine is in a unique position to enhance the evidence-based movement by encouraging increased attention to treatment mechanisms and to knowledge translation.

CONCLUSIONS

Clinicians, researchers, and, importantly, the public will benefit from the evidence-based movement by having a health care system that is built on solid grounds of evidence in determining which treatments should constitute the standard of care. A full partnership between clinicians and researchers is called for to generate the practical, rigorous evidence base needed to take behavioral health treatments to the next level of scientific support and implementation.

Age-dependent differences in short-term outcome after surgical or endovascular treatment of unruptured intracranial aneurysms in the United States, 1996-2000

INTRODUCTION

Unruptured intracranial aneurysm patients are frequently eligible for both open surgery ("clipping") and endovascular repair ("coiling"). We compared short-term end points (mortality, discharge disposition, complications, length of stay, and charges) for clipping and coiling in a nationally representative discharge database.

METHODS

We conducted a retrospective cohort study using Nationwide Inpatient Sample data from 1996 to 2000. Multivariate logistic regression analyses adjusted for age, sex, race, payer status, geographic region, presenting signs and symptoms, admission type and source, procedure timing, hospital caseload, and possible clustering of outcomes within hospitals. The results were confirmed by performing propensity score analysis.

RESULTS

A total of 3498 patients had clipping, and 421 underwent coiling. Clipped patients were slightly younger (P < 0.001). Medical comorbidity was similar between the groups. More clipped patients had urgent or emergency admissions (P = 0.02). More coiling procedures were performed on hospital Day 1 (P = 0.007). When only death and discharge to long-term care were counted as adverse outcomes, there was no significant difference between clipping and coiling. On the basis of a four-level discharge status outcome scale (dead, long-term care, short-term rehabilitation, or discharge to home), coiled patients had a significantly better discharge disposition (odds ratio, 2.1; P < 0.001). With regard to patient age, most of the difference in discharge disposition was in patients older than 65 years of age. The degree of difference between treatments increased from 1996 to 2000. Neurological complications were coded twice as frequently in clipped patients as in coiled patients (P = 0.002). Length of stay was longer (5 d versus 2 d, P < 0.001) and charges were higher ($21,800 versus $13,200, P = 0.007) for clipped patients than for coiled patients.

CONCLUSION

There was no significant difference in mortality rates or discharge to long-term facilities after clipping or coiling of unruptured aneurysms. When discharge to short-term rehabilitation was counted as an adverse event, coiled patients had significantly better outcomes than clipped patients at the time of hospital discharge, but most of the coiling advantage was concentrated in patients older than 65 years of age. Even in older patients, long-term end points-including long-term functional status in patients discharged to rehabilitation and efficacy in preventing hemorrhage-will be critical in determining the best treatment option for patients with unruptured aneurysms.

[Effectiveness of liberal vs. conservative episiotomy in vaginal delivery with reference to preventing urinary and fecal incontinence: a systematic review]

Episiotomy is the most common surgical intervention in the world. In Europe the rate of episiotomy is approximately 30% (23). Reasons for this intervention are the reduction of risk for tears and incontinence. To assess the effects of restricted episiotomy in the prevention of urinary and faecal incontinence. Medline search for 1990-7/2002, Cochrane Library (Issue 2, 2002), GEROLIT and SOMED and the Internet. RCTs analysing restrictive or non-restrictive episiotomy were included if they had comprehensive randomisation, follow-up and exclusion of selection bias. Cohort studies were assessed to evaluate the risk of developing faecal incontinence. If possible, data were pooled. Included were all pregnant women with vaginal delivery. Intervention/exposition: Restrictive vs. liberal episiotomy (median, lateral or mediolateral). Incontinence rate (urine and stool) 3 months and 3 years post partum. All included randomised controlled studies met the criteria above, one randomised controlled study used blinded assessment of outcome parameter. Lots of follow-up was 33% (after 3 years). Cohort studies partly were retrospective. 2 randomised controlled studies measuring urinary incontinence were included. The rate for episiotomy was 60% in the intervention group with liberal episiotomy and 27% in the restricted group. No difference could be found in groups measuring urinary incontinence (RR 0.98, 95% CI 0.83-1.20). Only two included cohort studies measured the effect of episiotomy on faecal incontinence. The chance of developing faecal incontinence in association with episiotomy was more than threefold (OR = 3.64, 95% CI 2.15-6.14). Restrictive episiotomy neither effects the development of urinary incontinence of post partum women (RR 0.98 95%, CI 0.83-1.20) three months and three years after vaginal delivery, nor the risk for trauma. Women without episiotomy suffer significantly less from faecal incontinence (OR = 3.6). Further investigation is required to measure the effect of no intervention versus liberal episiotomy.

Is patient travel distance associated with survival on phase II clinical trials in oncology?

BACKGROUND

Prior research has suggested that patients who travel out of their neighborhood for elective care from specialized medical centers may have better outcomes than local patients with the same illnesses who are treated at the same centers. We hypothesized that this phenomenon, often called "referral bias" or "distance bias," may also be evident in curative-intent cancer trials at specialized cancer centers.

METHODS

We evaluated associations between overall survival and progression-free survival and the distance from the patient residence to the treating institution for 110 patients treated on one of four phase II curative-intent chemoradiotherapy protocols for locoregionally advanced squamous cell cancer of the head and neck conducted at the University of Chicago over 7 years.

RESULTS

Using Cox regression that adjusted for standard patient-level disease and demographic factors and neighborhood-level economic factors, we found a positive association between the distance patients traveled from their residence to the treatment center and survival. Patients who lived more than 15 miles from the treating institution had only one-third the hazard of death of those living closer (hazard ratio [HR] = 0.32, 95% confidence interval [CI] = 0.12 to 0.84). Moreover, with every 10 miles that a patient traveled for care, the hazard of death decreased by 3.2% (HR = 0.97, 95% CI = 0.94 to 0.99). Similar results were obtained for progression-free survival.

CONCLUSION

Results of phase II curative-intent clinical trials in oncology that are conducted at specialized cancer centers may be confounded by patient travel distance, which captures prognostic significance beyond cancer stage, performance status, and wealth. More work is needed to determine what unmeasured factors travel distance is mediating.