Enterprise wide computerisation in hospitals not only helps in minimising treatment errors by eliminating mistakes made while prescribing medicines, but also helps in reducing the overall costs of the care provided.
Dr. Karanvir Singh
Consultant Surgeon & Head,
Sir Ganga Ram Hospital, New Delhi
Adverse incidents in hospitals compromising patient safety cost the governments billions of dollars each year. Indian statistics are not that easily available. This does not include the actual value of loss of life, when it occurs. It has been the aim of every ‘medical organisation’ to improve its safety record. Human beings make mistakes because the systems, tasks and processes they work in are poorly designed. The ideal solution is to improve processes and ensure adherence to the extent that no care provider, in any situation, can cause a patient related adverse incident. Enterprise wide computerisation can help to a certain extent in achieving this aim. In fact, there are many scenarios where enterprise wide computerisation is probably the only possible way of avoiding adverse incidents.
The most immediate barrier to widespread adoption of computerisation is cost. According to a study by RAND Health, the U.S. healthcare system could save more than USD 81 billion annually, reduce adverse healthcare events, and improve the quality of care if health information technology is widely adopted. It then makes sense to implement such technology early.
The most common and deadly treatment errors occur when physicians make mistakes prescribing medicines. Clinicians are most likely to make the following errors:
Prescribe drugs that interact with other medicines the patient is already taking, or with food the patient commonly eats.
Give patients a drug they are allergic to.
Give an incorrect dose or use an inappropriate frequency.
There are problems with manual medication order processing that are obviated with electronic ordering. Illegible handwriting, use of abbreviations (drug name, dosage, drug form, etc.), insufficient knowledge of drug-drug interactions, inadequate information on patient allergies and duplicate orders are some of these.
Once a patient’s allergy to a certain drug has been enquired into and documented by the care providers, it becomes the responsibility of the hospital to ensure that a drug, or related drugs, are never inadvertently administered. This applies not just to this particular admission episode but also to any subsequent ones. Mistakes do occur and it is not humanly possible to remember this allergy during subsequent admissions unless the patient informs this each time. A good computerised drug order entry system will display alerts whenever a care provider orders a drug to which the patient has been recorded to be allergic. This will be across episodes, i.e. even if the allergy documentation was done during a previous episode years ago. Drug class allergy alerts can warn in case a drug belonging to another class with allergy cross reactivity is given.
Other common alerts are for metal implants (avoid MRI), pacemakers (avoid electrocautery) or pregnancy (avoid ionising radiation and many drugs). While it may be difficult for hospital information systems (HIS) to manage alerts in such cases, a prominent display of such alerts in the patient record can help prevent adverse incidents. It should be possible to predefine end dates for such alerts, e.g. the LMP date in case of pregnancy, after which the alert will be automatically removed from the patient’s records.
Drug to Drug Interactions
With increasing specialisation, there are a larger number of doctors looking after a patient during the inpatient stay. Often each doctor prescribes medicines without full awareness of how this will interact with the other medications prescribed by other doctors. A good drug to drug interaction system will throw up an alert if it detects an interaction. Drug to drug interaction systems are somewhat complex to set-up and the efficacy of this system will depend on the pains that have been taken to configure this. No single off-the-shelf database is currently available in India that can be simply plugged into the HIS to provide this functionality.
A drug to drug interaction warning system will be more effective if it is doctors who are performing the drug prescribing directly in the HIS rather than nurses, as commonly happens, since they are more likely to respond to such warnings.
Drug Monographs and Dosing Alerts
Availability of full drug prescribing information at the time of ordering in the HIS is useful for the doctors. This itself can prevent, to a large extent, adverse drug interactions and dosing errors. Many systems are able to alert the care provider if the prescribed dosage is outside acceptable range although paediatric dosing alerts are found in fewer systems.
Batch Recall of Medicines from Wards
It often happens that a particular batch of medicines is found to be defective, sometimes in another hospital. With a HIS maintaining a track of all locations where that batch is currently stocked, it is possible to recall stock of that batch immediately. In such situations drug controllers often require details of all the patients to which that batch of medicine was administered. While this is not possible in a manual system, this should be easily possible in a hospital that has a HIS.
Patient sample mix ups are something that a good hospital policy set-up should be able to avoid. For instance, it should be mandatory to label the sample container before collecting the sample. There are other situations where a sample mix up can occur, e.g. while reading an illegible label or while transferring a sample to a machine.
A HIS can print out a label that is legible and is bar coded. All labels should also have enough human readable information on it to clearly identify the patient, specimen to be collected and the laboratory it is to be sent to. Further authentication can be done by requiring double scanning at the actual time of sample collection, one of the sample barcode and one of the patient ID barcode on the patient’s wrist.
A mix up at the time of loading the sample container in the autoanalyser machines can be avoided by using bidirectional interfacing of the machines where the sample container is directly plugged into the machine which is able to read the barcode and directly communicate with the HIS to get information of the tests to be performed and then send the test results back. Sometimes laboratories manually split the sample into different tubes when the samples are to be loaded into more than one machine. This negates the very advantage of bidirectional communication. The hospital policy should be able to avoid this by insisting on separate collections where such a sample splitting is anticipated.
Currently, there are no strong patient confidentiality laws in India but they could arrive very soon. A HIS can have an impact on patient confidentiality in many ways.
The main aim of a good HIS is to allow doctors to have immediate availability to all information about the patient that would be useful in treatment. A necessary trade off is in confidentiality. It is difficult to devise policies that would restrict patient data on the grounds of confidentiality and at the same time allow even junior medical and paramedical staff access to patient information. Even labelling a patient as a VIP patient, which allows restricted access of records to senior physicians only in our scenario, was counterproductive since it reduced the quality of medical care that essential medical staff could provide to these patients. It is mandatory by law to keep the results of certain tests confidential, e.g. tests for HIV, and a good HIS should allow this. For instance in our system it is possible to not only mark the result as confidential, i.e. visible only within the laboratory performing the test, but also mark the very request for this test as confidential so that care providers cannot see if such a test had even been ordered for. We have had problems in this with the nurses placing orders for the test again and again, not being able to see that it has already been ordered.
Clinical Guidelines for Disease Management
Inappropriate / incorrect medical care, not according to internationally accepted care guidelines can leave to patient morbidity and mortality and open the hospital to lawsuits. A good HIS with an electronic medical record that incorporates clinical pathways will act as a ‘mentor’ to the care provider and ensure that patients receive accepted care. These systems will outline the day to day activities to be performed in the form of a check list, for example they could prompt for anticoagulant therapy prior to any operation and check X-ray and mobilization on day one after an orthopaedic operation and so on.
Improved medical care by timely reminders of tests to be performed and vaccination schedules, though not strictly in the realm of patient safety do help improve the delivery of medical care. There are systems available in India that allow SMS alerts to be sent to patients as well as care providers.
The Flip Side
There are certain types of errors that have started occurring after the introduction of computerisation in hospitals. These constitute the flip side of computerisation. Some such errors represent the electronic version of “traditional” errors, such as a patient receiving the wrong drug dosage due to a human error. Now, in the computerized world, this can occur simply because the user clicked slightly away from the actual place on the screen. Another error is if the wrong patient is selected and orders placed. It is recommended that patient name, gender, and date of birth or age must be visible throughout the ordering process. Shortcut or default dosage selections can get ordered even for elderly, paediatric or underweight patients, resulting in toxic doses. Care providers may be led to a false sense of security that when technology suggests a course of action, it is always right. But on the whole, overall patient safety is likely to improve with installation of a computerised hospital information system.