Dianostic Error

Share |
The clinic dutifully followed the instructions for "E" codes, workman compensation codes. It was a simple fracture of a finger. When the response, with payment, came back, the diagnosis listed was gonorrhea resulting from a fall from aircraft.

Who is to say how this error occurred. Was it the insurance clerk exercising humor, a mistaken entry, a change in code numbers by the commission or an error in theirs or our computer.

Unfortunately, the above is not the only source of spurious diagnosis. In many cases the insurance clerk faced with, looking up the proper ICDA code from the clinician's notes, picks the one that justifies the laboratory and treatment ordered. She does so in  order to receive remuneration -- her responsibility. Even when the clinician lists the appropriate ICDA, the clerk may likely change it in order to met the criteria for payment. This manipulation of diagnostic code might be considered fraud. On the other hand, the clinic provided an honest service, rendered a diagnosis that may not exactly fit the codes and criteria. The clerk is fulfilling her duty to the clinic, the coding system and the patient. The clinic is therefore resolving ambiguities in good faith. However, the coded diagnosis may not accurately reflect the diagnosis for the patient.

Moreover, many clinicians are reluctant to record any diagnosis that the insurance company can call a preexisting condition. A noble concern, but one that erodes the acknowledgement of  early risk or the identification of incipient disease.

When one adds to these sources of erroneous diagnosis and the all to frequent missed or wrong diagnosis, there results a database of demographics and diagnosis that is corrupt from the start. A system of confirmed diagnosis might help.

Will the ACA or the EHR eliminate these sources of error? Probably not, the patient confidentiality issue remains. Even though the insurance company may no longer be able to deny preexisting conditions, the patient information is none the less in their database. The draconian rules limiting access to patient data makes it difficult for clinicians to coordinate care or the patient to access his or her information; whereas the system makes the information readily available to insurance companies and government. If Equifax wants to know if you have HIV, they are going to find out.

Lastly, how do you know that the diagnosis is correct even at best. Autopsy results find as much as 60% missed or wrong diagnosis. That is not to say that missed or wrong diagnosis occurs 60% of the time, but among patients who die, the missed or wrong diagnosis may be critical information. Shamefully, autopsy is a thing of the past. Blame it on families, fears of litigation, cost or hospital inconvenience, a revenue issue.  Autopsy is the final analysis, but who will pay for it?

Undiagnosed Hypertension

Share
|With 36 million estimated cases of uncontrolled hypertension, 14 million of whom are unaware and undiagnosed, there exists a target rich environment for the life saving diagnosis of hypertension.

Why should this be when most of these people visit clinics and doctor's offices, but remain unaware? A small number suffer the rather high threshold some doctors set for the diagnostic criteria for hypertension. The majority, however, represent a negligence in both the taking of blood pressures and the recording of the results stemming largely from cost cutting and sheer boredom.

Very few assistants know how to take a blood pressure. Rather than teaching them the right way and training them to a level of competency, accuracy and consistency, they are turned loose either with superficial knowledge or with an automated blood pressure cuff and recording instrument that may or may not be accurate and is still dependent on the assistant's ability to use the instrument correctly. If a clinician wants to treat more patients that are appreciative and save lives, the clinician would be well advised to take blood pressures his or herself.

Mechanized BP apparatuses notoriously give spurious results even the most expensive in the wrong hands. Moreover, what good is an automated cuff in the hands of a physician; he or she might rather use a mercury manometer and be certain. Today's hurried schedules with highly discounted reimbursements lead to a focus on a single problem. Ancillary abnormal findings get overlooked or discounted as probably in error. Sad but true, this happens and what more relevant condition can there be than the early diagnosis of hypertension along with the appropriate workup for underlying causes.

A proper BP reading requires an appropriate size cuff, a mercury manometer, and a stethoscope. The cuff should go first to the right arm in order not to miss Coarctation of the Aorta. The patient must fully extended the arm and raise it to chest level. The assistant must palpate a strong pulse placing the bowel of the stethoscope over the pulse while pumping the pressure to a safe level above the audible pulse, approximately 200 mm. Open the valve only a little so that the pressure drops slowly. Record the pressure for the first audible pulse, a rather distinct and abrupt point, easy to determine. Continue listening as the pressure falls. Record the first abrupt weakening change in sound and continue listening until no sound is heard recording that number as well. In some patients, there will be only one clear disappearance of sound while in others there are two changes to consider for the diastolic pressure, record them both. With only borderline BP, take another reading and do so for both arms. Record all readings such as 120/80-70 RA sitting. Include the position of the patient as well as the arm or leg. Always take multiple readings if pressure is above 120/70 and if above 140/90 take BP in one of the legs using a large cuff. Relegate BPs to only well known and trusted nurses and check even that during your physical exam. Take the patient’s pulse yourself as well; both make interesting conversation during your exam and lead to relevant system review questions.

The JAMA article below speaks of treatment protocols, but the diagnosis and workup come first. A treatment protocol may unerringly select the right treatment for the statistically average American patient but miss completely the appropriate treatment for the patient sitting before you. Many other conditions influence the choice of medications for treatment.

http://jama.jamanetwork.com/article.aspx?articleID=1778410&utm_source=Silverchair%20Information%20Systems&utm_medium=email&utm_campaign=JAMA%3AOnlineFirst11%2F14%2F2013

 

What Diagnosis?

Share | In the early nineteen sixties Larry Weed promoted the problem oriented medical record (POMR) in order to focus physicians on all of a patient's problems rather than focusing mainly on the chief complaint. No doubt the problem oriented approach was a systems improvement resulting in an improvement in medical care. Indexing all of the problems on a list with dates and resolutions made better sense of the patient record. The POMR was widely accepted but usually implemented in a mixture of the traditional source oriented medical record and the system promoted by Larry Weed. A medical school may title the teaching as POMR and then proceed to teach source oriented history taking with a chief complaint and a problem list.

Some complain that the POMR tends to focus on treating problems whilst ignoring diagnosis. Indeed, medical records and reimbursement documents force a qualifying diagnosis, but anyone wrestling with the ICDA diagnostic codes knows that a diagnosis can be written at various levels, four digits or five digits. For example are we treating cough, community acquired pneumonia or pneumonia due to a specific organism? The list is  long and reflects historical diagnoses based on gross findings. A physical diagnosis without regard to the bio-molecular underpinnings may be based on the problem list. For example: gastritis, hypertension, pneumonia, colitis, arthritis, arrhythmia, etc. 

The qualifying diagnosis for the guidelines may not accurately reflect the true underlying condition.

GBD omissions

Share | What about TB and Malaria? Christopher Dye and Mario Raviglione point out that GBD did not include TB. They call for a broader range of disease determinants for future studies.[1] One might also point out the omission of Malaria. TB and Malaria are the two most widespread and lethal diseases. TB especially could once again take hold in the US due to the concurrence of HIV and the emergence of TDR-TB, totally drug resistant organisms.

"This proposal goes beyond TB: for many causes of ill health, an unidentified risk is a missed opportunity."
_________________
Nature 502, 10 Oct 2013 Perspective, Weigh all TB Risks, Tuberculosis Outlook, S 13

Lim, s. s. et al Lancet 380, 2224-2260 (2012)

Global Burden of Disease (GBD) ranking

Share
|

JAMA published a seminal investigation documenting the Health of America compared to other developed countries broken down by disease, risk factors, morbidity and mortality. The massive undertaking involved hundreds of collaborators in thirty-four countries and some highly sophisticated statistics comparing the burden of disease. The report here includes the developed countries participating in the Organization for Economic Co-operation and Development.  The State of US Health, 1990-2010, Burden of Disease, Injuries and Risk Factors appears in the August 14 issue of JAMA.[1] If you have any interest in the shortcomings of our health care system or the health of our own human resource, you should read this landmark study.

Harvey Fienberg’s editorial on page 585 observes the decline in the US standing among developed nations.  Herein, “results for the United States are presented in detail for the first time.”  The statistics utilized in the study were extremely complex and all inclusive. However, Fienberg points out the fact that socioeconomic s was not included as a risk factor in this study.  The collaborators agree. The difficulty was in equating socioeconomics across the many cultures in the thirty-four nations that took part in this study. The editorial further reminds us of the well-established fact that socioeconomic status relates strongly with mortality[2] Fienberg further observes that the assessment for the US as a whole does not account for significant regional differences. On the plus side, the editorial suggests that the framework for assessing the burden of disease is scalable and applicable to states, counties and municipalities.[3]
The statistical terms used in this study include:  Years of Life Lost due to premature mortality (YLL), Years Lived with Disability (YLD), Disability Adjusted Life Years (DALY), which combines YLL and YLD and, Healthy Life Expectancy (HALE). The article compares these attributes in the thirty-four countries between 1990 and 2010. This monumental study goes beyond all previous reports by including risk factors for disease. In 2010 in the US the big eight YLL were in order: Ischemic heart disease, lung cancer, stroke, COPD, road injury, self-harm, diabetes and cirrhosis. The first eight YLD, however, were:  Low back pain, major depression, other musculo-skeletal, neck pain, anxiety disorder, COPD, disorders resulting from drug use, and diabetes.

The diseases causing premature mortality, YLL, differed dramatically from those causing morbidity and disability, YLD. However, the risk factors underlying the leading causes for both YLL and YLD diseases were almost the same. The list of risk factors included: Dietary, tobacco, HBP, high BMI, physical inactivity, high serum glucose, ambient particulate pollution, alcohol, drug use and high cholesterol.

The figure 4 illustration on page 604 reflects the US’s low YLL ranking compared with the thirty three other countries. The US is 7th from the bottom following the Check Republic and Chile in overall ranking. Our ranking by the raw longevity score, infant mortality and perinatal mortality reported elsewhere[4] are even worse.[5] The US comes in 38th in both longevity and infant mortality. The perinatal mortality rates are even worse.[6]

Figure 4.

Rank of Age-Standardized YLL Rates Relative to the 34 OECD Countries in 2010
Numbers in cells indicate the ranks of each country for each cause, with 1 representing the best-performing country. Countries are sorted on the basis of age-standardized all-cause years of life lost (YLLs) for 2010. Diseases and injuries contributing to YLLs are ordered by the difference between the US rate and the lowest rate in the Organization for Economic Co-operation and Development (OECD) countries for each cause. Colors indicate whether the age-standardized YLL rate for the country is significantly lower (green), indistinguishable (yellow), or higher (red) from the mean age-standardized YLL rate across the OECD countries. HIV indicates human immunodeficiency virus.[7]

---

[1] JAMA,2013;310(6):591-608. Doi:10.1001/jama.2013.13805
     Lim, S. S. et al Lancet 380, 2224-2260 (2012)
[2] National Research Council; Institute of Medicine. US Health in International Perspective: Shorter Lives Poorer Health. National Academies Press; 2013
[3] Katz B. Bradley J. the Metropolitan Revolution. Brookings Institution; 2013
[4]  http://www.enagic.com/enagic_life.php
[5] http://data.worldbank.org/indicator/SP.DYN.IMRT.IN
[6] http://whqlibdoc.who.int/publications/2006/9241563206_eng.pdf
[7] Figure 4 reproduced with permission JAMA RightsLink and Copyright Clearance Center Aug 14 JAMA, Christopher Murray et al; Copyright © 2013, American Medical Association

 

Virus vs Bacterial Respiratory Infection

Share |Science Translational Medicine reports an RT-PCR that can distinguish a viral pneumonia from a bacterial pneumonia. Christopher Woods and Geoffrey Ginsburg at Duke claim that the assay monitors  human genes that react differently to viral disease than to a bacterial infection. 
Based on a trial of 102 patients with fever and respiratory symptoms, the test showed 94 percent sensitivity and 89 percent specificity.
PCR might be an expensive test at the clinical level. It would be nice to run such tests as a routine if the small clinical lab can have the technology without involving big pharma patents.
With the new terminology for basilar rales, namely "crackles," one might expect the new physician to be confused over the identification of the subtle left lower lobe sounds which sound nothing like crackles. The PCR might help and more so for upper respiratory infections. Professor Kemp once told me as a resident on pediatric rotation that I was treating an upper lobe pneumonia with a lower lobe antibiotic. It won't hurt to look at the gram stain either.
--------------
Sci. Transl. Med. 5 203ra126 (2013)

Malaria Vaccine

Share
|

Science, Aug 8, 2013, DOI: 10.1126/science.124800, reports a new highly successful Malaria vaccine. Stephen Hoffman's intravenous (IV) vaccine consists of live attenuated plasmodium falciparum sporozoites -- labeled pfspz.

In a group of volunteers given four or five intravenous injections of the vaccine: 0/6 of the volunteers receiving five injections contracted malaria when later inoculated (p=0.018); 3/9 of the volunteers given only four injections (p=0.028); and 5/6 of the untreated volunteers contracted malaria. Despite the small population, the results were significant.

Prior experiments had shown that multiple mosquito bites by infected but radiated Mosquitos induced immunity. Hoffman took the process forward in employing numbers of radiated sporozoites. Sub-cutaneous injections did not work well, but the intravenous injections did.

Logistics in Africa may pose a problem in that the vaccine requires delivery in liquid Nitrogen containers. The IV dosing may not be so difficult because the volume is quite small, but getting patients back for four more injections likely will be. The authors suggest including the malaria vaccine along with other routinely delivered frozen veterinary products as a feasible avenue. Go.nature/mae5tu has a helpful summary on the background development of the vaccine.

Prior malaria vaccination attempts proved only slightly effective. Malaria remains, despite slow improvement, the number one killer worldwide with deaths estimated by WHO to be between 490,000 and 836,000 in 2010 and cases worldwide between 154 million and 289 million.

Global warming may once again result in Malaria's spread to northern regions, as p. vivax did in the first half of the twentieth century as far north as Archangel, Russia. (Lat: 64.5333)

The emergence of an effective vaccine comes at a good time. It remains to be seen if the vaccine works on other than the p. falciparum variety and on young children.