Computerised clinical decision support systems and absolute improvements in care: meta-analysis of controlled clinical trialsBMJ 2020; 370 doi: https://doi.org/10.1136/bmj.m3216 (Published 17 September 2020) Cite this as: BMJ 2020;370:m3216
All rapid responses
Re: Computerised clinical decision support systems and absolute improvements in care: meta-analysis of controlled clinical trials
In the era of Computerization and electronics, the convergence of Computer science and medicine has revolutionized the healthcare delivery system. Computerized decision support systems brought the paradigm shift in healthcare provisions, is a critical component of precision medicine and quality patient care by conversion of data to knowledge and sharing this knowledge for decision making and improvement of care. The integration, security and utilization of data which otherwise was challenging is made possible by utilization of a digital format. The system meets the expectations of clinicians by virtue of its flexibility, individuality and reliability to significantly improve performance, enabling them to attain the higher level of quality care. Commonly administered through electronic medical records and other computerized clinical workflows, it has revolutionized the complex decision making process of the clinicians since its inception in the 1980s. The utmost systemic tool, now facilitated by ever rising global adoption of electronic medical records with advanced capabilities circumvents new dimensions to error free healthcare .
Apart from all of the success stories, the system is not free from pitfalls and potential harms. The evidence-based recommendations for minimizing risk in the system design, implementation, evaluation, and maintenance is the order of the day.  The implementation of a generic, multi-purpose System having several disease covering entities with an evidence based application and functionality in primary as well as super speciality care is desirable; however, it is in its early phase in most health care systems . The effect on users, patient outcome and cost still eludes it from perfection.
In the case of stand-alone systems, the fragmented workflow disrupts information outflow from source space since it was designed without human information processing and behaviours in mind . Clinicians are often presented with excessive and unimportant alerts, predisposing them to alert fatigue. Disruptive alerts should be limited to more life-threatening or consequential contraindications.  Before the introduction of this system, healthcare providers, pharmacists, and nurses relied exclusively on double checking orders. This system creates the impression that verifying the accuracy of an order is unnecessary or automatic . This is an important myth to dispel. It is potentially problematic as the user has less independence and will be less equipped for that task should they switch to an environment without the system, thereby impacting user skill. Lack of technological proficiency can be disadvantageous for engagement with this system, which is dependent on computer literacy.
System maintenance poses a big challenge as maintenance of knowledge-base and its rules is unable to keep pace with the fast changing nature of medical practice and clinical guidelines. Even the most advanced healthcare institutions report difficulty keeping their systems up to date as the knowledge keeps changing. The operational impact of poor data quality and incorrect content is turning out to be another hiccup.  Medication and problem lists can be problematic, if not updated or used appropriately. Transportability is so difficult to achieve, and it suffers from interoperability issues as well.  Financial viability is on a par in a mounting significance to a vast majority of users. The system augments healthcare providers' decisions and patient care tasks, and with changing time actively and ubiquitously supports delivery of quality care.
At the same time, we must stay vigilant for potential downfalls of the system, ranging from simply not working and wasting resources to fatiguing providers and compromising quality of patient care. Advancement should include an input panel for formatting problems and available data and their related most suited algorithms of analysis to ensure that both healthcare providers and patients have the best fit method to inspect the situation and assist clinical decision making in a specific situation.
1. Sutton, R.T., Pincock, D., Baumgart, D.C. et al. An overview of clinical decision support systems: benefits, risks, and strategies for success. npj Digit. Med. 3, 17 (2020). https://doi.org/10.1038/s41746-020-0221-y
2. Helena Varonen, Tiina Kortteisto, Minna Kaila,et al, What may help or hinder the implementation of computerized decision support systems (CDSSs): a focus group study with physicians. Family Practice; 2008; 25: 3, 162–167, https://doi.org/10.1093/fampra/cmn020
3. Kilsdonk, E., Peute, L. W., Riezebos, R. J., Kremer, L. C. et al, Uncovering healthcare practitioners’ information processing using the thinkaloud method: From paper-based guideline to clinical decision support system. Int. J. Med. Inf. 86, 10–19 (2016).
4. Ash, J. S., Sittig, D. F., Campbell, E. M., Guappone, K. P., et al. Some unintended consequences of clinical decision support systems. AMIA AnnuSymp. Proc. AMIA Symp. AMIA Symp. 2007, 26–30 (2007).
5. Sujansky, W. Heterogeneous database integration in biomedicine. J. Biomed. Inform. 34, 285–298 (2001).
Competing interests: No competing interests
Reading the article by Kwan et al  one has to be disappointed about the outcome: the poor benefit perceived from clinical decision support systems (CDSS) embedded within the electronic health records (EHR). And it has to be questioned if the study reflects the truth, as it contradicts the perception one could have as a clinician.
In the article introduction, comparing the USA and UK it is pointed “EHR adoption has been slower in the UK5” . I rather prefer the statement that the USA should consider “studying the experience of a nation [the UK] farther along the road” . The fact is primary care in the UK has had 100% EHR coverage for well over a decade. It is not the same for secondary care, and the reference used by Kwan et al pointed to that particular case .
The use of EHR, the benefits from moving from paper records, get diluted when this meta-analysis implies there are little benefit from them, and concludes that CDSS need improving to reduce the threat of alert fatigue.
In my view, the main threat preventing a better outcome is human factor variability. The EHR can only respond to data input in the right format, and unfortunately, at least in primary care, there is considerable variability on the use of EHR . Failing to consider the human factor, this systematic review relates to Smith and Pell's wonderful paper on parachute use to prevent death . The conclusion is wrong, the focus suggested is wrong. Artificial intelligence decision based on human intelligence entries cannot be better because of the variable and often minimal interest by clinicians to code all information properly, which is not helped by the fact that there is lack of training and lack of a simple to use nomenclature to input the relevant data into the EHR.
1. Kwan L, Lo L, Ferguson J, Goldberg H, Diaz-Martinez JP, Tomlinson G, Grimshaw JM, and Shojania KG. Computerised clinical decision support systems and absolute improvements in care: meta-analysis of controlled clinical trials. BMJ 2020;370:m3216 http://dx.doi.org/10.1136/bmj.m3216
2. Payne TH, Detmer DE, Wyatt JC, Buchan IE. National-scale clinical information exchange in the United Kingdom: lessons for the United States. Journal of the American medical informatics association. 2011;18(1):91-8.
3. Sheikh A, Jha A, Cresswell K, Greaves F, Bates DW. Adoption of electronic health records in UK hospitals: lessons from the USA. Lancet 2014;384:8-9. doi:10.1016/S0140-6736(14)61099-0.
4. Millares Martin P, Sbaffi L (2020). Electronic Health Records (EHR) and Problem Lists in Leeds UK. Variability of General Practitioners’ views. Health Informatics Journal, vol. 26 (3), 1841-1865. https://doi.org/10.1177/1460458219895184
5. Smith GC, Pell JP. Parachute use to prevent death and major trauma related to gravitational challenge: systematic review of randomised controlled trials. Bmj. 2003;327(7429):1459-61.
Competing interests: No competing interests