Clinical Decision Support Need for Standardization
Submitted by Dr. Chandrashekhar Bhoopalam
Clinical Decision Support System (CDSS) is interactive software that assists physicians in decision-making about their patients. The system utilise data from pharmacy, laboratory, radiology, and other patient monitoring systems to help physicians in enhancing patient care. Statistics show an increase in the number of medical institutions adopting CDSS in pursuit of reducing errors, to improve the nursing documentation and improving patient outcomes. This paper discusses errors arising from the use of CDSS and ways of preventing them.
A recent research by Gary and colleagues found out that CDSS improved performance of healthcare practitioners by 64%. A previous research by Kawamoto et al depicted an improvement of 68%. The two researches noted that an automatically generated system provided better delivery of care than one that is user initiated (NCBI 2005). Despite the huge costs of implementing a CDSS in healthcare centres and the vast improvement on patients care, errors are inevitable in usage. First, a CDSS is only as strong as its knowledge base. An ancient knowledge base will issue outdated diagnosis and medications that can have adverse effects on the patient. Medical field is one that evolves with time and a decision based on a database only two years old might have such diverse effects.
To correct and avoid such an error, healthcare institutions using CDSS should ensure proper maintenance of their databases. Such databases should be adaptive to provide continuous updating and reflect the most current information (NCBI 2005). Secondly, according to AHRQ some CDSS are complex and require data entry of the patient. This is useful when a patient presents confusing uncertain diagnosis. An incorrect initial data entered means incorrect results. This happens in the non-automated systems and the answer to this is to automate the CDSS and ignore the human error component (AHQR Web)
After clinical acceptance of the system, clinicians face some challenges in some areas. According to stead and sittig, while many health centres have implemented systems successfully, others have remarkably failed to incur significant delays and cost overruns (stead and Sittig 108-122). One of them is the range of biological systems that require utilization of all the relevant data. This may vary from family history, medical history, genetics, trends of the problem in terms of its prevalence, and existing published data on this condition on medical effectiveness. The technicality involved causes a disruption in the normal workflow. Thirdly, some errors in the systems may result from abbreviations used in the programming of the CDSS. For instance, an abbreviation on the drug prescriptions like every other day, every day, mg, mcg, and the use of decimals may have different interpretations (AAOS). An error occurs when a nurse giving the medication misinterprets it. This error is avoidable by inclusion of the entire medical team in developing the system to agree on which abbreviations to use.
Another disadvantage is that most CDSS are standalone and do not include the existing health systems. Clinicians consume lots of time in re-entering data in the CDSS that is available in the current health plan. The system may generate an enormous quantity of alerts, which may aggravate the clinician and cause him to ignore them. Critical alerts can be missed in the process and make the case to get the wrong medication and eventual death. According to Arch internal medical journal, in an investigation that identified 1111 cases of errors, only 64.4% of the cases could be prevented by a CDSS (Bobb A, Gleason K, Husch M et al 785-791). Although the use of CDSS helps avoid unnecessary errors, it is inevitable to clear all types of errors.
Different CDSSs serve different purposes, which makes it difficult to assess the system effectiveness. According to Open clinic website, CDSS cause interoperability problem on local, regional, and national levels that affect communication (Open clinic web). A system with irregular communication makes it difficult to assess. Lack of proper research over the years has hindered inclination to focus only on functional decision. According to AHRQ, although some studies exist on CDS, few are randomised controlled trials. On the other hand, researchers concentrated on effects of CDS on the process of care and not on the outcomes or the structure. Additionally, researchers concentrate on clinician’s decision making other than ignoring diagnostic programmes, which have had limited use in practical settings (AHRQ Web)
AHRQ. “Clinical Decision Support Systems” Web, 29 May 2012, WWW.healthit.ahrq.gov/images/jun09cdsreview/09_0069_ef.html>
AHRQ. “Clinical Decision Support Systems” Web, 29 May 2012, www.ahrq.gov/clinic/ptsafety/chap53.htm>
AOOS. “Prevention of Medical Errors.” 2012. Web, 29 May 2012 www.aaos.org/about/papers/advistmt/1026.asp>
Bobb A, Gleason K, Husch M et al. “The epidemiology of prescribing errors: the potential impact of computerized prescribe order entry” Journal of Arch Intern Med. 2004 12;164(7):785-92.
NCBI. “What Makes a Good Clinical Decision Support System”2005; 330 (7494) 740-741. Print.
Open Clinical. “Computer Entry Order Entry Systems” Web, 29 May 2012, www.openclinical.org/cpoe. Sittig DF, Stead WWW. “Computer-based physician order entry: The state of the art.” Journal of Am Med Inform Assoc. 1994; 1: 108–123