HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Gottlieb, R. H. Search for Related Content PubMed PubMed Citation Articles by Gottlieb, R. H. Social Bookmarking                      What's this?

Imaging for Whom: Patient or Physician?

¹ Roswell Park Cancer Institute, Elm and Carlton Sts., Buffalo, NY 14263.

Received March 18, 2005; accepted after revision June 7, 2005.

 
Address correspondence to R. H. Gottlieb.

Abstract

Top Abstract Introduction The Role of Imaging The Role of Technology The Role of the... Information Technology as a... References

 
OBJECTIVE. Radiologic imaging examinations are being ordered beyond the margin of medical necessity. Radiologists can assess the value of imaging in a variety of clinical situations by gathering data regarding test ordering patterns and their effects on patient outcomes.

CONCLUSION. Emerging information technologies have the potential to facilitate the collection of data and permit the dissemination of appropriate guidelines to limit the number of unnecessary and possibly harmful examinations.

Introduction

Top Abstract Introduction The Role of Imaging The Role of Technology The Role of the... Information Technology as a... References

 
Radiology services accounted for only 3.5% of total personal health care spending in 1990, but expenses for high technology imaging have had one of the most rapid growth rates of any type of health care in recent years [1]. Between 1998 and 2001, high technology imaging, including MRI, CT, sonography, and nuclear medicine studies increased for Medicare enrollees between 7% and 16% per year whereas conventional radiology increased 1% per year [2]. In 1960, the United States devoted 5.0% of its gross national product to health expenditures. By the mid 1990s, that figure had risen to 13.5%, more than $1 trillion, and continues to increase rapidly [1, 3].

Assuming radiology services account for the same percentage of health care expenditures currently as in 1990, most likely an underestimate, diagnostic imaging and imaging-guided interventions add billions of dollars directly to medical expenses and potentially motivate additional financial outlays for treatments based on the findings of the tests or complications related to the interventions. Savings are possible if the imaging results in patients being spared more expensive treatments, shortens the course, or lessens the severity of illnesses. It is clear that diagnostic imaging in medicine, although obviously not the sole determinant of increased health care expenditures and the quality of health in this country, plays a very important role with regard to these issues.

The Role of Imaging

Top Abstract Introduction The Role of Imaging The Role of Technology The Role of the... Information Technology as a... References

 
Very little objective effort has been devoted to assessing the role of imaging in various medical scenarios despite the large potential impact imaging has on both the quality and cost of medical care. Factors promoting the use of imaging have been previously biased in terms of benefiting manufacturers, avoiding physician malpractice risk, and promoting academic careers. An additional factor is that we radiologists have passively accepted the rapid rise in ordering of high-technology imaging studies by our physician colleagues.

Increased testing using expensive imaging (primarily CT), despite the reduction in reimbursement on a per test basis from previous fee-for-service levels, has coincided with a substantial rise in job positions and incomes for radiologists [4]. Third-party payers have, to date, served as the only consistent gatekeepers of our technology, and their motivations are often directed more toward cutting costs than promoting patient health.

Control of imaging in our society is a multidimensional issue, and some of the determinants of the use of radiology services are beyond the control of radiologists. Radiologists certainly cannot change social imbalances leading to the large number of Americans without health insurance that drives an increased use of emergency services. Increased use of these relatively inefficient services is often associated with higher costs of treatment and testing for medical conditions that could be dealt with more inexpensively and effectively if these same patients had vigilant outpatient follow-up by dedicated practitioners.

Radiologists cannot control self-referral by our physician colleagues in cardiology, vascular surgery, or obstetrics and gynecology. However, the high use of radiologic imaging by practitioners, driven by fear of malpractice suits or ignorance regarding the optimal imaging evaluation of a patient's problem, can be altered by our profession. Radiologists need to become more proactive in the evaluation of imaging in specific patient scenarios and disseminate this information effectively to those making the decisions regarding test ordering. Recently presented data show that only a small minority of malpractice claims arise from the lack of ordering additional imaging studies (Baker S, Harkinsoon S, presented at the Radiological Society of North America meeting, December 2004).

New imaging technologies (MDCT, PET) and the expanding uses for existing imaging tests for which patient harm by the tests is perceived as minimal or nonexistent are leading to a rapid increase in imaging. CT use is the most striking example of this trend. Studies have recorded a four-to fivefold increase in CT examinations for select clinical situations [5, 6]. CT scanners are more numerous and scanning time, including image reconstruction, is far faster than in the early 1990s, removing the impediments to scanning that existed previously.

CT has largely replaced the excretory urogram in detecting urinary calculi, clinical evaluation in assessing patients with suspected appendicitis, and nuclear medicine imaging in patients with suspected pulmonary emboli. Superficially, these innovations seem to advance the level of medical care; however, one must consider that imaging technologies are currently used on a mass scale within our society. What might benefit select individuals who really present diagnostic dilemmas to our clinical colleagues may not be the case for many others who are being scanned because of fear of malpractice claims or misconceptions about what specific tests should be offered in select clinical scenarios.

CT is considered by many clinical practitioners as a "window into the body," [7] a way to reduce diagnostic uncertainty, and therefore anxiety, about missing a potential serious diagnosis or to minimize their risk of being sued. Although imaging can increase physician diagnostic certainty, and this may be of value in facilitating patient care, two recent outcome studies in patients with lower-back pain found imaging can improve physician confidence regarding patient diagnoses without any actual effect on patient treatment or clinical outcome [8, 9]. What is frequently left out of the equation when weighing whether to scan is the potential harm to the patients, including added radiation exposure (especially to younger individuals), reactions to IV contrast material, or downstream effects of false-positive and false-negative interpretations. Misinterpretation of CT may become more pronounced for new, innovative scanning techniques as these techniques move out of academic medical centers and into community practice.

The Role of Technology

Top Abstract Introduction The Role of Imaging The Role of Technology The Role of the... Information Technology as a... References

 
Technology diffusion has been a well-studied phenomenon in medicine, and previously developed concepts can be applied to the dissemination of radiologic imaging techniques. A number of factors promote the diffusion of new techniques or technologies in radiology and have been discussed in the radiology literature [10]. These factors operate at the level of the individual, who makes the decision to order the test or promote usage of the test, or at the organizational level, with groups supporting purchasing a device, selling the device, or making rules and regulations regarding the use of the device.

Early communication regarding the preliminary effectiveness of an imaging test for a particular clinical situation at scientific meetings or through advertising by the manufacturer may result in premature widespread adoption of the test for an ever-expanding list of indications and patient groups who may not benefit from the results of the test. Once an imaging technique is incorporated into medical practice to evaluate a particular clinical problem, reversing the decision to do so is difficult. Controlled studies evaluating the condition with and without the test, once the test becomes an accepted part of practice, are difficult to perform because of ethical and financial constraints. It is therefore exceedingly important to conduct high-quality initial technology evaluations before diffusion occurs and tests become established in diagnostic pathways in an ever-enlarging circle of patients who may not benefit from the new technique.

The speed and perceived noninvasiveness of the newer imaging techniques, most notably CT, facilitate a lack of critical evaluation of whether they are being used beyond the margin—the margin being the point where the use of medical resources is accompanied by an equivalent medical benefit to patients. As high as 30% of radiology studies may be unnecessary [11, 12]. Unrestricted use of imaging may actually result in patient harm, although that may not always be apparent to referring physicians, radiologists, third-party payers, and patients. Several examples illustrate the concept of scanning beyond the margin of medical necessity.

CT has rapidly replaced the excretory urogram in the detection of urinary tract calculi. Initially touted benefits regarding the technology in this clinical scenario included not requiring IV contrast material, the speed and ready availability of the test, and the detection of other unsuspected causes of the patient's symptoms such as appendicitis or gynecologic masses. The test was rapidly embraced in the emergency department setting, where diagnostic uncertainty is high, and there is a pressing need for rapid patient turnover.

One recent outcome study found diagnostic imaging for patients with suspected calculi increased by 26%, many of whom were under the age of 30 [5]. The same study found no significant change in patient outcomes compared with a similar time period when the excretory urogram was the predominant imaging study. Hospital admissions and return visits to the emergency department were the same for both time periods. An unsuspected outcome was an increase in urologic interventions (stone extractions, lithotripsy) for stones discovered by CT considered too large or far from the ureterovesical junction to pass. This group experienced a small number of procedure-related complications. Many of the patients ( 50%) were repeat stone formers; some of these patients had multiple CTs over the course of the same calendar year for suspected stones. Study recommendations for curtailing scanning beyond the margin included no testing for many repeat stone formers who could be sent out of the emergency department with symptomatic relief and instructed to strain their urine.

In a similar manner, CT has rapidly replaced ventilation-perfusion nuclear medicine scanning for suspected pulmonary emboli. Cited advantages of CT over nuclear medicine imaging include speed of diagnosis, personnel availability, fewer indeterminate studies, and detection of other abnormalities besides pulmonary emboli. As with CT for stone disease, diagnostic testing for suspected pulmonary emboli has increased substantially [6]. The ACR appropriateness criteria recommend an initial chest radiograph before either CT or nuclear medicine imaging, which can reveal a cardiac or pulmonary cause for hypoxemia [13]. Nuclear medicine imaging is still recommended for patients with a normal chest radiograph as an effective test to exclude pulmonary emboli.

Scanning beyond the margin is becoming widespread for many patients in whom the likelihood of a pulmonary embolus is low. CT is often the first test ordered. The switch from nuclear medicine scanning to CT has not been proven to reduce the morbidity and mortality of pulmonary emboli in the population despite increased imaging to rule out the condition. A simple blood test evaluating fibrin split products has not been integrated into the diagnostic testing algorithm. This test provides an inexpensive, noninvasive way to identify patients likely to have an embolus and exclude other patients from further workup [14].

CT has recently been incorporated into the workup of patients with suspected appendicitis. Outcome studies have validated a more accurate identification of patients with appendicitis with CT over clinical evaluation alone [15, 16]. However, CT may not improve on clinical evaluation in certain patient subgroups, including children and women of reproductive age [17, 18]. Surgery has been reduced in patients without appendicitis with no increase in perforations in patients discharged from emergency departments. However, as with the two examples cited previously, CT is being used beyond the margin of medical necessity to exclude appendicitis for an ever-increasing number of patients with abdominal pain. Many of these patients do not have a high pretest probability for appendicitis, lacking an elevated WBC, fever, or rebound tenderness on physical examination. A higher number of false-positives can be expected in this more liberally selected patient population for the examination. Individuals being scanned for appendicitis are young and the radiation burden on this population will not be known for years.

Efforts to curtail imaging beyond the margin of medical necessity have been limited and difficult to enforce on a case-by-case basis. With high-technology imaging, most notably CT, now readily available and perceived as largely noninvasive, there has been no sustained effort to reduce the volume of imaging. Little attention has been given to the down-side of imaging on a massive scale and beyond what is medically necessary in terms of the resulting false-positive and false-negative test results, population radiation burden, reactions to IV contrast material, patient anxiety and discomfort regarding being scanned, and the expense of the additional procedures.

The Role of the Radiologist

Top Abstract Introduction The Role of Imaging The Role of Technology The Role of the... Information Technology as a... References

 
Radiologists have largely abdicated their role as consultants to our physician colleagues in the majority of expensive imaging studies ordered. Initial studies showed direct consultation between radiologists and clinical colleagues resulted in a substantial reduction in ordering expensive imaging procedures such as CT and sonography [19, 20]. More recent evaluation of the effect of radiologic consultation regarding expensive imaging studies on inpatients at two academic centers failed to confirm any substantial effect. There was little impact on the number or types of imaging studies ordered [21, 22]. However, these studies were conducted at teaching hospitals where specialists are often involved in diagnostic workups and are often knowledgeable about appropriate imaging in their areas of expertise. Generalists ordering outpatient examinations or imaging in the emergency department setting were not evaluated in these studies.

Attempts to affect physicians' behavior in ordering radiology testing have been perceived as time-consuming, potentially antagonistic, or both and have been reserved for only the most absurd or dangerous examinations or procedures. Unfortunately, most imaging beyond the margin involves large-scale ordering of examinations that have no real benefit or perhaps may even cause harm, though not readily apparent, to those patients who undergo them.

There has been no systematic effort, other than by third-party payers, to restrict the marked overuse of radiologic imaging because of little perceived benefit from doing so. Radiologists do not, for the most part, see any advantage to spending large amounts of time consulting with referring physicians on a case-by-case basis that in the majority of occasions is futile; may generate ill will in future dealings with the same physicians; and would, if successful, potentially curtail imaging volume and therefore their incomes.

Referring physicians see no benefit in restricting their overordering of imaging studies because, in most cases, they think they are doing the right thing to gain certainty regarding their patients' diagnoses or are protecting themselves from a possible malpractice suit. Private industry (manufacturers, pharmaceutical companies) clearly would be adversely affected by restricting imaging beyond the margin through loss of sales because fewer imaging devices and pharmaceuticals would be purchased if imaging volume decreased significantly.

We as radiologists have a moral responsibility to change the current status quo of imaging beyond the margin. If we are provided information that overuse of imaging in many routine scenarios could hurt our patients, then we are obligated to act to change the situation leading to this behavior on the part of our physician colleagues. How can we accomplish this? A three-pronged approach must be used. First, we must rapidly gather information regarding the effect of imaging studies on patient outcomes. Second, we must widely disseminate this information to our physician colleagues to affect their test ordering behavior in a time-efficient, nonthreatening manner. Third, we must gain support for these activities by relevant outside agencies including private third-party payers, the government, and hospitals.

Searching for the truth regarding the value of imaging in specific clinical scenarios is the first step toward a more rational use of radiology services. This process is a difficult one and revolves around obtaining valid information rapidly and effectively about how our imaging studies affect patient care and clinical outcomes. The most valid tool to accomplish this is the randomized clinical trial. However, this method is time-consuming, expensive, and may not be realistic or moral when asking patients to consent to trials of a potentially effective imaging study versus doing nothing. Because of the inherent long periods of data acquisition, the technologies evaluated by the randomized trial method may be outdated by the time conclusions are reached.

Retrospective studies are less accurate than randomized trials because there are many potential biases that can affect conclusions. They also, by their very nature, may have outdated results when released because of innovations or improvements in existing technologies that may change test performance characteristics rapidly. Retrospective studies often rely on tracking information through medical record searches, which is a tedious, time-consuming process.

Information Technology as a Tool

Top Abstract Introduction The Role of Imaging The Role of Technology The Role of the... Information Technology as a... References

 
Fortunately, information technology has evolved as a potential tool to help acquire the data needed to answer questions regarding the validity of imaging in specific patient clinical scenarios. A joint effort is needed between radiologists and clinical colleagues to establish order entry systems that are easy to use and provide reliable data to the radiologists interpreting imaging studies. Novel interfaces are required linking the patient medical record to requests for imaging examinations.

Key data points need to be included in the information provided to a radiology record such as patient allergies, prior surgeries, and relevant structured indications for the examination. The indications need to be coded in a format that can later be searched when reviewing the relevance of specific imaging studies in select clinical circumstances. Radiology reports also need to be structured with findings and impressions coded to allow later searches regarding the validity of the information provided in guiding clinical treatments and correlating with future pathology and surgical findings or subsequent patient outcomes [23, 24]. Side benefits of structured reports include increased consistency between radiologists and improved clarity of the information provided to referring physicians.

Clearly defined indications for examinations through the order entry component of the system can be linked to well-understood terminology in structured radiology reports describing patient conditions. The usefulness of imaging studies in specific patient clinical presentations can be evaluated by the percentage of times radiology interpretations positively affect patient outcomes or correctly predict surgical or pathologic findings. The potential exists for online, real-time research because the information needed to assess test relevance can be captured at the time the examinations are performed and reviewed quickly by mining the existing database.

Such a system requires close cooperation between radiologists, referring physicians, software programmers, administrators, and legal advisors. The order entry component must be simple to use, possibly directly linked to notes entered in the medical record through novel interfaces with our referring physicians. Well-designed templates in combination with voice recognition may be useful. Direct links to other information system data bases (laboratory values, past medical records, allergy history, other relevant contraindications for testing) need to be established.

Creating a mobile platform may be useful for physicians unable to sit at a console because they are on rounds or in a busy emergency department. Structured reporting systems, currently in their infancy [24], must be further developed to be easy to use for large-volume radiology practices and yield more standardized, searchable information provided in the findings and the impressions. Annotated images may be useful, with an edited set of images provided to referring physicians with links to the images on the PACS. The terms used to describe the findings on the images must be carefully chosen and provided on menus to ensure consistency.

Rapid mining of the information gained from such a system may allow the development of guidelines either at the local level (hospital, community) or nationally after pooling the data from multiple sites. These guidelines can be placed in the order entry component of the system as prompts to be used as needed by referring physicians [25, 26]. Individually consulting with referring physicians regarding the usefulness of every test is inordinately time consuming for reducing imaging beyond the margin, but system prompts require no such time investment and can, in a nonthreatening manner, be used by referring physicians.

Use of computerized order entry systems, with system prompts, has previously been shown to reduce ordering errors of pharmaceuticals [27]. If guided by system prompts, which would be based on evidence-based practice, some degree of protection from malpractice suits may be gained as well as allowing for more rational test ordering. Outcome projects could test this hypothesis by comparing physicians using or not using system prompts. Some degree of compliance could be gained by tracking when physicians use and do not use system prompts and following patient outcomes in both situations.

The future is now in the world of information technology. Ease of use of order entry and reporting systems in radiology has been a major stumbling block but should be a surmountable problem, especially considering how far gaming technologies have evolved. Money and time are obviously needed to allow us to have some fun using the system. Programmers paired with referring physicians and radiologists need to devise creative interfaces on both the front end (order entry) and back end (reports) to develop a system that is universally acceptable.

Support for such a system must come from hospitals, equipment vendors, third-party payers, and our physician colleagues. Radiologists would be unable to develop such a system in isolation. Linking the front-end information (the reasons for the tests) to the back-end information (the results of the tests) can allow online outcomes research regarding test relevance in specific clinical areas. Tremendous information can be gained to enable radiologists to meaningfully consult through such a system with our physician colleagues to appropriately guide them in their test ordering, stem the rising tide of ordering beyond the margin, and avoid harm to the patients we are sworn to protect.

References

Top Abstract Introduction The Role of Imaging The Role of Technology The Role of the... Information Technology as a... References

 

1. Margulis AR, Sunshine JH. Radiology at the turn of the millennium. Radiology 2000;214 : 15-23[Abstract/Free Full Text]
2. Bhargavan M, Sunshine JH. Utilization of radiology services in the United States: levels and trends in modalities, regions, and populations. Radiology 2005;234 : 824-832[Abstract/Free Full Text]
3. Levit K, Cowan C, Bradley B, Stiller J, Sensenig A, Lazenby H. National health expenditures in 1997: more slow growth. Health Aff (Millwood) 1998; 17:99 -110[CrossRef][Medline]
4. Forman HP, Kamin DS, Covey AM, Sunshine JH. Changes in the market for diagnostic radiologists as measured through a help wanted index. AJR 2000; 174:933 -938[Abstract/Free Full Text]
5. Gottlieb RH, La TC, Erturk EN, et al. CT in detecting urinary tract calculi: influence on patient imaging and clinical outcomes. Radiology 2002;225 : 441-449[Abstract/Free Full Text]
6. Prologo JD, Gilkeson RC, Diaz M, Asaad J. CT pulmonary angiography: a comparative analysis of the utilization patterns in emergency department and hospitalized patients between 1988 and 2003. AJR2004; 183:1093 -1096[Abstract/Free Full Text]
7. Gunderman RB. The medical community's changing vision of the patient: the importance of radiology. Radiology2005; 234:339 -342[Free Full Text]
8. Gillan MG, Gilbert FJ, Andrew JE, et al. Influence of imaging on clinical decision making in the treatment of lower back pain. Radiology 2001;220 : 393-399[Abstract/Free Full Text]
9. Jarvik JG, Hollingworth W, Martin B, et al. Rapid magnetic resonance imaging vs radiographs for patients with low back pain: a randomized controlled trial. JAMA 2003;289 : 2810-2818[Abstract/Free Full Text]
10. Ramsey SD, Hillman AL, Renshaw LR, Kimberly JR, Pauly MV, Schwartz JS. How important is the scientific literature in guiding clinical decisions? the case of magnetic resonance imaging. Int J Technol Assess Health Care 1993; 9:253 -262[Medline]
11. Cascade PN, Webster EW, Kazerooni EA. Ineffective use of radiology: the hidden cost. AJR 1998;170 : 561-564[Free Full Text]
12. Brown RF, Shaver JW, Lamel DA. The selection of patients for x-ray examinations. Rockville, MD: Bureau of Radiological Health, U.S. Food and Drug Administration publication 80-8104;1980
13. American College of Radiology. ACR appropriateness criteria 2000. Radiology 2000;215 [suppl]
14. Burkill, GJ, Bell JR, Chinn RJ, et al. The use of a D-dimer assay in patients undergoing CT pulmonary angiography for suspected pulmonary embolus. Clin Radiol 2002;57 : 41-46[CrossRef][Medline]
15. Brandt MM, Wahl WL. Liberal use of CT scanning helps to diagnose appendicitis in adults. Am Surg 2003;69 : 727-731[Medline]
16. Torbati SS, Guss DA. Impact of helical computed tomography on the outcomes of emergency department patients with suspected appendicitis. Acad Emerg Med 2003;10 : 823-829[CrossRef][Medline]
17. Kosloske AM, Love CL, Rohrer JE, Goldthorn JF, Lacey SR. The diagnosis of appendicitis in children: outcomes of a strategy based on pediatric surgical evaluation. Pediatrics2004; 113:29 -34[Abstract/Free Full Text]
18. Flum DR, Morris A, Koepsell T, Dellinger EP. Has misdiagnosis of appendicitis decreased over time? A population-based analysis. JAMA 2001; 286:1748 -1753[Abstract/Free Full Text]
19. Baker SR, Rosenberg ZS, Adel H. The operation of a ward-based radiology consultation service. Radiology1984; 152:331 -333[Abstract/Free Full Text]
20. Khorasani R, Silverman SG, Meyer JE, Gibson M, Weissman BN, Seltzer SE. Design and implementation of a new radiology consultation service in a teaching hospital. AJR 1994;163 : 457-459[Abstract/Free Full Text]
21. Gottlieb RH, Hollenberg GM, Fultz PJ, Rubens DJ. Radiologic consultation: effect on inpatient diagnostic imaging evaluation in a teaching hospital. Acad Radiol 1997;4 : 217-221[CrossRef][Medline]
22. Bree RL, Kazerooni EA, Katz SJ. Effect of mandatory radiology consultation on inpatient imaging use: a randomized controlled trial. JAMA 1996; 276:1599 -1600[CrossRef][Medline]
23. Huang Y, Lowe HJ, Hersh WR. A pilot study of contextual UMLS indexing to improve the precision of concept-based representation in XML-structured clinical radiology reports. J Am Med Inform Assoc 2003; 10:580 -587[Abstract/Free Full Text]
24. Langlotz CP. Automatic structuring of radiology reports: harbinger of a second information revolution in radiology. Radiology 2002;224 : 5-7[Free Full Text]
25. Khorasani R. Computerized physician order entry and decision support: improving the quality of care. RadioGraphics2001; 21:1015 -1018[Free Full Text]
26. Tjahjono D, Kahn CE. Promoting the online use of radiology appropriateness criteria. RadioGraphics1999; 19:1673 -1681[Abstract/Free Full Text]
27. Bates DW, Leape LL, Cullen DJ, et al. Effect of computerized physician order entry and a team intervention on prevention of serious medication errors. JAMA 1998;280 : 1311-1316[Abstract/Free Full Text]

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				B Hofmann and K B Lysdahl Moral principles and medical practice: the role of patient autonomy in the extensive use of radiological services J. Med. Ethics, June 1, 2008; 34(6): 446 - 449. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Gottlieb, R. H. Search for Related Content PubMed PubMed Citation Articles by Gottlieb, R. H. Social Bookmarking                      What's this?