Los Angeles County Grand Jury • 2006-2007

Emilie Anselmo George Buckley Nola Burnett

Published: December 06, 2006 353 pages Consolidated Report
View Original PDF

Findings 11 findings

F1 Page 43
DHS central staff Directors have established spheres of influence when working with the Hospitals, but they lack formal authority relationships and audit functions Current relationships are built on trust and collegiality and judicious use of staff authority. The authority has not encompassed all the oversight provisions included in the Job Specification for the Director, Pharmacy Services. The Director, Pharmacy Services, reports to the Departmental Chief Medical Officer. The Director, Pharmacy Services, participates in and provides staff support for major committees, including the DHS Core Pharmacy and Therapeutics Committee, the DHS Pharmacy Director’s Committee, and the DHS Pharmacy Information Systems Committee. The incumbent has been effective in funding and generating agreement about implementation of new technology. This assignment of responsibility and chain of command through the Chief Medical Officer appropriately reflects the need for central direction for the Hospitals. Nevertheless, the Hospitals do not share the full reports of all Medication Errors, which are essential for the Director to understand overall Pharmacy Performance and needed interventions. In addition, there is no formally developed Audit Schedule to review conformance with Departmental policies associated with ordering, distributing, and administering medication.
F2 Page 44
Investments in technology can result in significant reductions in medication error risks. Automated medication dispensing cabinets have proven themselves to be a good, if substantial, investment. Where they have been put into use, Hospital staff members report that errors have fallen, and the dispensing process has become more efficient. Similar results are projected for pharmacy medication carousels. Bar coding systems offer the prospect of a new level of patient medication safety that has heretofore been unavailable. A pre-requisite for the success of these systems is the implementation of standards for bar code systems across the industry. IV Smart Pumps also add a new level of medication error reduction by highlighting delivery rates or concentrations that fall outside standard parameters. Smart Pumps allow for programming of standard concentrations of IV medications, removing the burden of the unit nursing staff having to calculate flow rates, thereby eliminating an additional potential source of error. The implementation of new technology has rendered information technology staff ever more vital to Hospital operations. The medication use systems assessed here are highly dependent on contractors and internal staff members for installation, staff training, and system maintenance. The systems are expanding in both scope and complexity. DHS is aware of this issue. DHS’s Chief Information Officer (CIO) has prepared a business plan that outlines the resource needs he expects to face immediately and for several years into the future. Work remains to gain the funding needed to offer competitive salaries and deploy staff at the Hospitals.
F3 Page 45
Olive View Medical Center processes appear to be incompatible with realization of ADC benefits. Olive View’s description of its plans to install ADCs is not consistent with the approaches used by other Hospitals. Its current systems include pneumatic tube distribution of medications. Citing space limitations on the floors, Olive View staff members are designing plans that limit access to the ADCs, which limits its advantage of allowing nursing staff complete access to regular medications while maintaining controls over controlled substances. During the site visit to Olive View, it was not conclusively demonstrated that patients’ identities are always recorded on medical records before medical orders are written. There was some evidence that the patient’s last name was being written by hand on the initial order sets and then covered with a printed label. This is not a best practice, and it affords a clear opportunity for medication error, as it is possible for the order to refer to the wrong patient. When questioned about the practice, Olive View staff denied that this was the case. The denial, in fact, is more disturbing than the practice itself, since it suggests a defensiveness that could cover other errors or weaknesses. A single site visit should not be considered conclusive, and is not in itself sufficient grounds for action. It is, however, sufficient to support a review of current medication orders, dispensing, and administration to ensure that the Hospital is fully prepared to implement a more automated pharmacy system.
F4 Page 46
Electronic Health Records (EHRs) and Computerized Physician Order Entry, when properly implemented, are best practices to reduce medication error. Computerized Physician Order Entry (CPOE) has been successfully implemented in several regional hospitals. It is difficult and expensive to do. It requires extensive planning, a significant investment, and careful building of internal acceptance. Executive management must make the commitment to CPOE and consistently let people know that it will be implemented – especially when thorny issues arise or the system shows difficult growing pains. A pre-requisite for its success is the implementation of EHRs, which are currently being planned in DHR. CPOE, difficult as it is to implement, is worth the trouble. It eliminates a transcription step that generates errors. It eliminates illegible orders and unapproved abbreviations. Wrong dose and wrong drug errors are reduced. Data can be easily collected for evidence-based clinical paths, considered a best practice. Physician orders, including medication orders, can be transmitted to the responsible parties instantly. Medical records can be shared between facilities and programs, making it much faster and easier to assess patients accurately, and all but eliminating the “shopping” of medical services by patients. Productivity, quality of care, and patient safety all advance. DHS has been wise in the order of technology implementation. The technologies described in this report are well-proven innovations that impact patient safety. After the installation of Automated Pharmacy Systems, assignment of Clinical Pharmacists, acquisition of IV Smart Pumps and bar coding of medications are complete, EHRs and CPOE appear to be the next major steps in reducing the risk of medication errors.
F5 Page 47
Clinical Pharmacists are effective in reducing medication error risks. Clinical Pharmacists spend time as part of patient care teams, consulting with Physicians and the remainder of the team on the units and during rounds. They provide drug information, recommend dosage adjustments, respond to questions from nursing staff, and make suggestions regarding initiation of drug therapy. All of these activities relieve stress on the Physicians and other clinicians who must sometimes make rapid decisions about patient care in difficult circumstances. This improves the quality and speed of the decisions and reduces the likelihood of medication errors.
F6 Page 48
24-hour Pharmacist coverage is important to Patient Safety. A Pharmacist is needed on-site for in-patient care on a 24/7 basis, particularly for acute patients. If a Pharmacist is not available, a comparatively slow and cumbersome process of faxes and telephone calls to Pharmacists at home is required. A “night locker” must be maintained, under the control of a Senior Nurse. If a medication is needed that is not in the night locker, a Pharmacist must be called in from home to open the main pharmacy and dispense it. The primary disadvantage of the process is that it is not fast, and the circumstances that give rise to overnight medication orders are likely to demand prompt action. There are also evident weaknesses in the processes for the checking of medications against orders. Finally, tired Pharmacists are expected to perform their day shifts, even when their sleep is interrupted. In all professions, errors usually increase when people become sleep-deprived. Some smaller hospitals cannot support a 24-hour pharmacy. These hospitals are adopting the capabilities of automated prescription transmission, which involves scanning medication orders to a Pharmacist for verification and approval. The Pharmacist can then release the medication from the ADC at the smaller facility.
F7 Page 49
It is a common practice to conduct pre-employment drug testing among staff involved in the health care professions. DHS does not do this. Drug screening and random drug testing are common means of ensuring security and safety in a wide variety of sensitive business and government operations. A number of hospitals use them. Failure to conduct screening and testing exposes patients to a potential risk of not receiving prescribed medication (if an addicted employee substitutes a placebo for the particular drug), and increases the potential risk of medication errors when impaired staff members prescribe, approve, distribute, and administer medications. This is a known risk in an environment where staff members sometimes work long hours, and where there is ample knowledge of the “right” mix of drugs to take to stay alert.
F8 Page 50
A “Safe and Just” culture is essential to accurate reporting of errors – including medication errors – as well as to the rapid and effective response to error trends. A survey undertaken by DHS in December 2005 showed disturbingly high levels of concern among both Nurses and attending Physicians that errors would be held against them. Nurses also expressed high levels of concern that errors would be kept in their personnel files. Some of this concern may be traced to a long-standing policy in which managers close out reports of all types of medication errors by indicating that an employee was “counseled about the incident”. That is the first step in the County’s standard progressive discipline process. In that process, managers are advised to keep a permanent record of the counseling action for use as needed in ensuing sanctions for further errors. In April 2006, DHS engaged a contractor to deliver a training workshop, entitled “Patient Safety and the Just Culture”. It appears to have been intended to address the issue of a punitive culture by advocating a culture in which acknowledgement that errors are human and inevitable is balanced by a strong sense of professionalism and “zero tolerance” for medication errors. Before punitive actions are taken when a rule is broken or mistake is made, a Just Culture would consider: Whether the employee knowingly violated the rule (cid:104) Whether there was a compelling reason to violate the rule (cid:104) Whether the employee had a good faith but mistaken belief that the violation (cid:104) was justified or insignificant Whether the source of the error resided within the system or was behavioral. (cid:104) Based on answers to such questions, the response would vary from counseling to training to revising procedures. Perceptions of an unjust culture are crippling to any attempt to encourage the self-reporting of Category A, B and C errors, in which no measurable harm comes to a patient. If Physicians and Nurses fear retribution, they will be significantly less likely to self-report errors that have little measurable impact. DHS’s efforts to counter them are laudable, but it seems likely that they have not been sufficient. Further surveys were not available, but anecdotal evidence suggests that negative perceptions have proved to be persistent.
F9 Page 51
There is insufficient consistent, Department-wide medication error trend analysis. The identification of trends within overall medication error rates is key to reducing them. The Hospitals do some analysis now, although they do not all follow the same guidelines; and their analysis appears to be limited at times by a lack of formal training for some Hospital Risk Management staff members. For the past year-and-a-half, data have been gathered in a nationwide system run by the University HealthSystem Consortium, which permits sophisticated analysis of errors. There has been insufficient consistent reporting and analysis to take advantage of that capability. While some errors occur only within one Hospital, others may be related to overall policies, training regimens, or technology systems across the County. These error trends should be identified and corrected; this can only be done if errors are reported and analyzed consistently across the County. In addition, such a function will serve to protect the Hospitals against the possibility that a medication error trend may be missed at the Hospital level. Recently, DHS has developed a “Report Card” which tracks high level trend data. It is a good beginning, but needs additional development.
F10 Page 52
The number of errors in UHC-PSN Categories A-C is less than expected, and raises the possibility that they are not fully self-reported. A review of reported errors shows few reported errors in the A and B1 categories for all four Hospitals in the six months ending March 31, 2007. The total errors reported for “no patient harm” categories (A, B1, B2, C) were low, considering there were over 300,000 patient days at Harbor-UCLA, Olive View, and RLA, where many patients have multiple medications. Given the number of reported errors in other categories over the same period, this does not appear logical. It is understandable that staff members who make lower category errors are sometimes reluctant to report them. Even in a non-punitive atmosphere, it is easy to see that a staff member would be reluctant to report a minor mishap, such as picking up an improper medication from a shelf, noticing it immediately, and replacing it before gathering up a proper medication. This is how major trends are identified that lead to, for example, separating look alike/sound alike medications. For errors in Categories B1 and B2, the patient receives no improper medications. Self-reporting is the only source of data. Responsible staff noted that the system used to track medication errors and pharmacy interventions is very cumbersome to use, further discouraging reporting of errors. This possible underreporting is important, because identifying errors that do not result in patient harm can point the way to weaknesses in the system that pose the potential for more serious adverse events.
F11 Page 53
DHS does not have a regular practice of replicating the best practices implemented from one facility to another. As an example, Harbor-UCLA has implemented Structured Physician Order Forms for admissions and transfers requiring the use of block lettering to reduce cases of illegibility. These forms significantly reduce a broad range of medication order errors, including illegible orders, unapproved abbreviations, dosage errors, and “wrong drug” errors. The Structured Physician Order Forms are considered successful where they are now used. The balance of DHS facilities does not use block lettering on these forms. As another example, Olive View does not recognize the same medications as high risk as the other Hospitals. These are examples of processes that are low-cost and can be relatively easily implemented, as contrasted with implementation of ADCs and bar coding technology.

Recommendations 11

Conclusions 10