I

n 1990, New Day Professional Services had many requests to develop automated clinical systems for long-term care facilities. At that time, the facilities were mainly interested in automated care plans. Shortly thereafter, the Minimum Data Set (MDS) was developed and enforced by Health Care Financing Administration or HCFA (now the Centers for Medicare and Medicaid Services or CMS). However, automation of the MDS was not a requirement, and, therefore, facilities were not as anxious to have clinical processes automated as they were financial processes.
       The nurse consultants at New Day Professional Services developed a basic system that provided automation for care plans and the MDS. When implementing the system in nursing homes, there was much resistance of the clinical staff to use and accept an automated system to assist them in completing their clinical assessments, care plans, and documentation. At that time, many nurses did not even know how to use a mouse! Even as the owner of New Day Professional Services at that time, I did not trust computers or feel they could assist in rendering acceptable care.
       However, 13 years later, the scenario is quite different! Now, not only do we see virtually every facility utilizing automation, we see too many facilities using the computer as a "be all/end all" for clinical processes, many where professional input is necessary by interdisciplinary staff members who are familiar with the resident.
       With the MDS now being automated and information submitted nationally by all certified Medicare and Medicaid facilities, many challenges arise as a result of the automated clinical information. Not only does the information impact the assessment process but also provides the information whereby quality indicators (QIs) and quality measures (QMs) are determined. The automated information also provides information for survey agencies to review prior to conducting the survey, and any condition noted on the MDS is made available to the surveyors through review of various forms, such as the Resident Level Summary, which identifies any of the 24 QIs that are present on the resident.
       While New Day Professional Consultants certainly endorse and encourage automation, there are many areas where we have witnessed automation actually having a negative effect on resident care outcomes as well as the survey process. Following are some examples of areas that are often utilized in automated programs for MDS, care plans, and survey process.

Pitfalls That Can Occur When Facilities Totally Depend on Automated Clinical Systems
       Many automated programs provide "bells and whistles" for the MDS, care plans, and survey items that are intended to assist the professional. However, the professional must be sure they are assuring accuracy by reviewing the information provided before proceeding with saving and transmitting the documents. Following are some examples.
       Automated systems allow users to carry forward responses from previous MDS to current MDS. While this is a very nice feature, if the professional does not review and validate that the information remained the same, often inaccurate information is entered and submitted. This is most noted on the MDS items for the response on condition changes. Maybe the response was "improved" on the previous MDS but there has not been any change on the current one, thus the code remains as "improved."
       Survey deficiencies have been cited for inaccurate assessments when information, such as "pain," was brought forward and not changed by the user. No documentation was present on pain relief offered during the assessment period, which resulted in a deficiency being cited.
       Often times, deficiencies are cited because interventions on care plans are not carried out as care planned during the survey process. Most often, the user had used the suggested approaches that came with the automated care plan, which were not appropriate for the particular resident. For example, a problem/need for a resident on anticoagulant therapy might have an intervention to "use a soft toothbrush," and this was placed on a care plan on which the resident had full dentures. This is often referred to as "canned" care plans even though most automated systems have methods to individualize the care plans.
       Some automated systems will identify significant changes that should not be identified according to the federal criteria. For example, the system looks for two areas of decline or improvement in coding, which may have been present just during the assessment time period or was expected due to medication or condition, and the user performed the significant change based on the computer information rather than interdisciplinary judgment.
       When surveys take place, facilities are asked to provide surveyors with CMS required reports: Resident Roster Matrix and Census and Condition reports. Many automated programs produce these reports from information based on MDS responses. There are also some items requested on the reports that cannot be obtained from MDS responses. Quite often, the user does not review and correct these reports based on current conditions of the resident, which is what surveyors expect. When the facility uses the computer-generated report without making corrections, the report contains information from the last completed MDS and does not necessarily reflect current conditions of the resident. It would also not have information for residents who did not have a completed assessment yet.
       Another area is identifying and clarifying resident assessment protocols (RAPs). Often, facilities only use the forms provided by the automated program and do not document or follow the guidelines as outlined in the federal Resident Assessment Instrument (RAI) manual. For example, the automated program may list the triggers and items from the RAP key but do not lead the user in further assessing those areas that are in the guidelines but not on the MDS form.
       Often, care plans are not updated as conditions or interventions change on the automated care plan and are often not carried forward the next time the care plan is reviewed and placed in the computer.
       Frequently, the automated program identifies problems for the care plans from the MDS and RAPs and the users do not add those problems/needs that may be evident that are not included in these assessments.
       Most automated clinical programs offer much more than just the MDS and care plans. However, with the survey process and many other processes, the MDS and those functions most used and associated with the MDS (QIs, QMs, care plans, etc.) become crucial when determining facility outcomes and survey results.
       Facilities should investigate why clinicians do not conduct reviews or make corrections, as they should for the previously mentioned pitfalls. Does the automated program provide methods to make the corrections? Is the software intuitive and is it easy to make the corrections? Are users able to override decisions imposed by the software? Does the software encourage individualization of the clinical processes provided by the computer?
       No matter the number or greatness of all the automated "bells and whistles" provided by automation, human judgment must be exercised throughout the entire clinical process. Facilities and users must understand that while automation can suggest, remind, encourage, and provide much needed information and analysis, the final responsibility and accountability falls on the shoulders of the clinicians. Surveys and outcomes are not based on the automated program but on the identification and implementation of clinical processes that will assure quality of life and quality of care.