Old Topics‎ > ‎

mort06_2

2006-09-19_01 Code for Fryn's syndrome

(Chapter XVII)

[Question from Sue Walker, Australia:]

We have had a question about the coding for a death certificate of a 2 year old female with Fryn's Syndrome. There is no specific code for this congenital autosomal recessive disorder but we have found further information about the syndrome at http://health.enotes.com/genetic-disorders-encyclopedia/fryns-syndrome.

We think that maybe it should be coded to Q87.0 but wondered if anyone has any other thoughts?


[Comments:]

2006-09-26:

Canada /Patricia Wood/:
We have had several questions this year about the coding of congenital anomaly syndromes named for different people. I don't know what the intention of the Classification is for Chapter XVII. In the case of multiple congenital anomaly syndromes with abnormalities that each go to different codes, should we code to the predominant anomaly, or perhaps to the most lethal of them? Personally, I am inclined to code them to Q897 Multiple congenital malformation, not elsewhere classified, but I do realize that this would result in a very long list of different syndromes being dropped into this code, with no way to distinguish one from another.

I hope that our discussions about Sue's question will lead us to develop a strategy for the classification of multiple congenital anomalies because I think we will continue to encounter them in the future.

Germany /Orlando Özer/:
We would propose Q87.8 as the adequate code for the Fryn's Syndrome, since this syndrome is also characterized by abnormalities of the organs (amongst others). Thus Q87.8 seems more appropriate to us than Q87.0.

Jordan /Majed Asad/:
Here is my suggestion for the Fryn's Syndrome case:
1) Q89.7, or
2) Q87.8.

2006-09-19_02 Drug-related deaths - police report but no death certificate

(Chapter V, XVII)

[Question from Pavlos Pavlou, Cyprus:]

We are still uncertain about coding some drug related deaths and I would be grateful for your assistance.

We get information on these deaths from police reports. We also collaborate with the Cyprus Centre for Monitoring Drugs and Drug Addiction. They maintain a Special Register for these deaths. Medical or forensic certification is almost always insufficient or non-existent. We have to do the underlying and multiple cause coding without having a properly filled medical death certificate. The following is a typical example of information provided by the police.

A 29 year old man, a known, long-term drug addict, was found dead, in a hotel room. A needle and syringe was found by the bed. He had obvious injection sites on his arms. There was no evidence or suspicion of suicidal intent or homicide. Toxicology showed high levels of heroin in the blood.

In such a case we assume that the death certificate would have included the term "long-term heroin addiction" and therefore we would code the UCOD with an F11 code. If however, there is no clear information on long-term drug addiction, we code the UCOD as X42.

In the absence of proper death certification, is this the correct way of handling these cases?


[Comments:]

2006-09-26:

Canada /Patricia Wood/:
I think that there may be more than one answer to this question! Prior to 2006 deaths due to drug poisoning with mention of drug addiction were coded to the drug addiction, but effective 2006 they are coded to the drug poisoning. This is as a result of URC recommendation 0117 (approved in October 2002 for implementation in 2006). So if this death took place in 2005 or before, I think that the UC would be F112, but from 2006 on the UC would be X42 with a nature of injury code T401.

Germany /Orlando Özer/:
In this case we would propose to code the External Cause : X42, also because the deceased had high levels of heroin within his blood. If there was the information that the circumstances persist undetermined instead of further investigation, one could also consider to code Y12.

2006-10-03_01 Unascertained natural causes

(Chapter XVIII, XX)

[Question from Coleen Hill, Australia:]

We are asking to see how other countries are dealing with the following scenario and would appreciate any help:

The coroners official finding was a 32 years old female died of Unascertained Natural Causes. There were a number of findings at autopsy which may be of potential significance in the sequence of events leading to death - Obesity; Left Ventricular Hypertrophy; Mild asthma; Second child, 2 months post partum (no complications in pregnancy); possible myocarditis'

We coded as follows:
1(a) R98
1(b)
1(c)
1(d)
II. E669; I517; J459; O96; I514

We are uncertain what the underlying cause should be. We get many cases similar to this; sometimes it says unascertainable natural causes on a background of many different diseases such as listed above.


[Comments:]

2006-10-16:

Canada /Patricia Wood/:
Like Coleen, we often see certificates that say "unknown cause" (or something similar) in Part I and then a list of causes in Part II. By the strictest application of the rules, the tentative underlying cause selected from Part I would be reselected by Rule A and obesity, the first of the causes from Part II, selected. No further linkages or modifications apply so obesity would be the underlying cause of death.

I think that this is open to debate because it would seem quite contrary to the intention of the certifier. If he states that the underlying cause of death is unknown, should we really be arbitrarily selecting something first in a list of conditions mentioned in an autopsy report? In Canada, we have had this conversation often, especially in the coding of underlying cause of stillbirth where the certifier says the cause is "unknown" but we might code P015 because the fetus was one of a twin pregnancy.

I believe that Rule A is useful for modifying the underlying cause from certificates where the tentative underlying cause of death selected from Part I is ill-defined (as presented in the examples in Volume 2), but maybe not so useful when the cause in Part I is specified as "unknown". However, perhaps the first approach should be to explain the situation to the certifier and him to specify the underlying cause. Is that likely to be fruitful Coleen?

New Zealand /Christine Fowler/:
We experience the same problem in New Zealand. We are reluctant to code to R98 or R99 as underlying cause where significant conditions are documented on the autopsy report, even if the coroner's official finding states 'Unascertained natural causes' just because the mode of death is unknown.

Generally, we would code the first significant condition mentioned as the underlying cause. Our justification for doing this is Modification Rule A, i.e. if a medical certificate of causes of death had been written for the death with 'Unascertained cause' in Part 1 and specific conditions listed in Part 11 then, in accordance with Rule A, you would re-select the cause of death as if the ill-defined condition had not been reported. In the example Coleen has given this would mean Obesity would be selected as the underlying cause.

Sweden /Lars Age Johansson/:
I agree with Patricia and Christine that following Rule A is really the only suggestion that ICD gives on cases like these. Just to add a further uncertainty to the discussion: it seems from some Swedish cases that I have seen that what the forensic examiner really means to say is that s/he doesn't know what the direct cause of death was, or why the patient died at that particular moment. The underlying cause(s) may still be more or less clear. So perhaps Rule A isn't such a bad idea after all.

2006-10-03_02 Multiple site chordoma

(Chapter II)

[Question from Elisabeth Gantenbein, Switzerland:]

Once a month we discuss some cases in our coding team. Here is one of them, we couldn't decide which coding is the best or the right one:

A 60 years old man, died of a Chordoma of the epipharynx, clivus and sinus sphenoidalis. How do you code this rare diagnosis? Malignant or benign neoplasm or neoplasm of unknown behaviour? Neoplasm of bones or of epipharynx  (localisation) or without specification of site?


[Comments:]

2006-11-14:

Brazil /Ruy Laurenti/:
I agree with the comments made by Martti from the Nordic Centre. This is a very good case for discussion in the MRG.

2006-10-24:

Germany /Orlando Özer/:
Though a "chordoma" is rarely malignant, the ICD-O-3 declares it as a malignant neoplasm.

Other doctrines say that it mostly arises from the clivus-region. The other sites mentioned (epipharynx, sinus sphenoidalis) are not typical osseous structures and could be sites of metastases etc.., but that's speculative.

Further the mentioned sites do not correspond to the same organ system. We'd preferably code C41.0, but one could also consider C97 according to section 4.2.7 H) of the Rules.

Nordic Classification Centre /Martti Virtanen/:
From eMedicine (http://www.emedicine.com/med/topic2993.htm)

"Chordomas are rare neoplasms. As primary intracranial neoplasms, they only constitute 0.2% of all CNS tumors; however, they constitute 2-4% of all primary bone neoplasms. Chordomas generally occur in 3 locations, which are, in descending order of frequency, the sacrum, intracranially at the clivus, and along the spinal axis. Fifty percent of chordomas occur in the sacrum, and spinal axis chordomas are rare. Occasional parasellar and sellar examples have been described, and extraaxial sites have been reported in the literature."

The tumor grows slowly with subtle symptoms for a long time and it rarely sends metastases. Thus original site in this case is obviously "at clivus".

Since this is not really a neural tumor, the only option seems to be C41.0. The ICD-O code M9370/3 Chordoma is specific, but does not help us in this situation.

For ICD-revision work, the important question is, is the localization of the a tumor more important than type? For example the statistics of chordomas disappear totally in the current system. And this is obviously not the only such case. Could MRG discuss this matter?

2006-10-17:

Canada /Patricia Wood/:
The lead term "chordoma" is in the index with a cross-reference to Neoplasm, malignant ... but from there, it's not so easy! The site "clivus" is coded to C410, Malignant neoplasm of bone and I did read that chordomas may arise from the intraosseous remnants of the notochord. The other sites mentioned are not indexed as malignant neoplasms of bone, so I am in a bit of a quandary. Maybe some one has some oncology expertise and can help us out with this one?

2006-10-03_03 Severe asthma due to food allergy

(Chapter X, XX)

[Question from Elisabeth Gantenbein, Switzerland:]

Once a month we discuss some cases in our coding team. Here is one of them, we couldn't decide which coding is the best or the right one:

Another not so rare case is the one of a 20 years old man: he died of a acute severe asthma due to an anaphylactic reaction due to adverse food (pumpkin seeds)? We discussed if it's better to code the asthma or the exposure to the substance. What's your opinion? In case of the exposure, which code do you suggest? Is it correct to take accidental poisoning by...

X499?


[Comments:]

2006-10-17:

Canada /Patricia Wood/:
Although we haven't had "pumpkin seeds" mentioned as the allergen before, we have tackled coding anaphylactic reaction to a food as the underlying cause of death. The issue was brought to the Mortality Reference Group and here is an excerpt from the discussion:
"The nature-of-injury codes related to food allergies are better in ICD-10 than ICD-9; however, there is ambiguity about what external code to use.

The MRG believe that future revisions should consider addressing allergy and autoimmune diseases and immunological reactions in the disease chapters. For ICD-10, X58 (Exposure to other specified factors) is the appropriate code." I think that this death should be coded to X58 with a nature of injury code T780.

2006-10-17_01 ACME coding of surgery

(Rule GP, 4.2.6)

[Question from Pavlos Pavlou, Cyprus:]

When we have a certificate like the one below,
Ia) Cardiorespiratory arrest
Ib) Nephrectomy for Ca
Ic) Renal Ca
II) Weakness, paralysis of extremities, uraemia,
we code as UCOD: renal carcinoma, C64, in accordance with Rule 3, "an operation on a given organ should be considered a direct consequence of any surgical condition (such as malignant tumor or injury) of the same organ reported anywhere on the certificate".

We also have in mind note "4.2.6 Operations", in vol 2 (which does not apply in this case, because the condition is mentioned on the certificate).

"ACME 2006.3" codes this certificate to UCOD: Y836, unless we manually correct it at the "edit MICAR Reject file" stage. The ACME decision tables also lead to Y836. What is the correct code? Are we misinterpreting Rule 3 ? Are we using ACME incorrectly ?

I would be grateful for your explanations, which are always enlightening.


[Comments:]

2006-11-14:

Brazil /Ruy Laurenti/:
According to General Principle I code this case as Renal Ca-C64 It is not Rule 3 as applied by Pavlos. The sequence a due to b and b due to c is accepted.

Canada /Patricia Wood/:
Pavlos is right - in the case he presents, the underlying cause of death should certainly be the condition necessitating surgery.

It is my impression that MMDS identifies all records with mention of surgery as MICAR rejects requiring manual resolution by multiple causes classification staff before processing through the ACME portion of the system. I agree with Gunvor and Lars Age; if the NCHS instructions for classifying multiple causes of death (as contained in Instruction Manual 2b) are applied to this certificate, ACME correctly selects C64 as the underlying cause of death.

2006-10-24:

Germany /Orlando Özer/:
We would code the renal carcinoma C64. This case would not justify to code an External Cause. The condition which made the operation necessary is clearly given on Line Ic).

The ACME code does not seem to result from an incorrect use of ACME, but it is another indication of the insufficient handling-options of procedures by an Automated Coding System.

Norway /Gunvor Ostevold/:
I think this problem is about how to use ACME. When you use an ampersand according to how to code surgery, procedures and complications, &C64, ACME will code C64 as uc.

Sweden /Lars Age Johansson/:
I agree with Gunvor, the problem is caused by specific multiple cause coding requirements in ACME. If a procedure appears alone on a line in Part I, you have to assign a code from Chapter XX, most often in the range Y60-Y84. (As Orlando says, perhaps this isn't quite what the external cause codes in ICD were intended for, but those using ACME just have to accept ACME's coding conventions, I'm afraid.) As Gunvor points out, you also have to assign an ampersand (&) to the condition that necessitated the procedure. In this case, the codes would be
1a) I469
1b) Y836
1c) &C64

With this coding, ACME selects C64 as the underlying cause. There is much more to procedure coding in ACME than this, of course. Full documentation is available at http://www.cdc.gov/nchs/about/major/dvs/im.htm - check the "2b" manual.

2006-10-17_02 Pneumoconiosis due to wheat flour

(Chapter X)

[Question from Monica Pace, Italy:]

How would you code this certificate? (We are currently using 2005 updates)
Man, 78 years
a) Irreversible respiratory failure
b) Pneumoconiosis
c) Wheat meal; chronic obstructive pulmonary disease

We are discussing about two possibilities and would like to know your ideas on the following solutions:
Underlying cause is not influenced in any case.
a) J969
b) J670
c) J448
Or:
a) J969
b) J64
c) J670; J448


[Comments:]

2006-11-14:

Brazil /Ruy Laurenti/:
In this case I prefer to code I67.0 See exclusion note under J44.-

Canada /Patricia Wood/:
Well, I really don't know very much about farms in general, or wheat meal in particular, but it seems that certifier is saying that the pneumoconiosis is due to wheat meal, making J670 seem like a very good code for line b)!

However, I would code COPD as J449 rather than J448. Then I would be inclined to consider J670 to be a more specific description of the nature of the underlying cause of death (Rule D) and I would select it.

2006-10-24:

Germany /Orlando Özer/:
A first possibility is to ignore the terminus "wheat meal" or to regard it as a further explanation of COPD so:
a) J969
b) J64
c) J448 ...seem to be the right codes;

But in this case 'wheat meal' could also be interpreted by 'J67.0', which is slightly unexact. This leads to:
a) J969
b) J64
c) J670; J448

2006-10-17_03 SIDS/SUDI [Sudden unexpected death in infancy]

(Chapter XVIII)

[Question from Christine Fowler, New Zealand:]

We would like to ask if any other countries are experiencing the following change in reporting sudden infant deaths, and if so how do they code them.

Some coroners and pathologists in New Zealand are moving away from using the term Sudden Infant Death Syndrome (SIDS), in favour of Sudden Unexpected Death in Infancy (SUDI). I believe the reason is that the term Sudden Infant Death Syndrome might mistakenly be considered to be a specific condition that causes death, whereas in fact it just describes a set of circumstances in which an apparently healthy baby dies of unknown cause whilst sleeping.

Coroners and pathologists want to make this clear, hence the name change to Sudden Unexpected Death in Infancy . Autopsy is performed for all deaths reported as SIDS or SUDI in New Zealand, and the SUDI deaths occur in bed just as the SIDS deaths do. In some SUDI deaths where a baby was sharing a bed with parents or siblings the pathologist has stated that, whilst there was no direct evidence of overlying, the possibility that the baby suffocated due to overlying could not be excluded. In other cases swabs taken of various body sites have cultured bacteria but autopsy has shown no evidence of active inflammation/infection in those sites. In one case microscopic examination of one of the lungs of a 5 month old baby showed early acute bronchopneumonia, however, the pathologist stated "the amount of bronchopneumonia present is, in my opinion, insufficient to account for the death of this child" and then stated "the death of this child should be categorised as Sudden Unexpected Death in Infancy rather than Sudden Infant Death Syndrome".

Currently we are coding SUDI deaths to the SIDS code (R95). Do others think this is correct or would R98 or R99 be more appropriate? We are concerned that if we code SUDI deaths to R98/R99 we would be artificially lowering the rate of SIDS deaths and might give a false impression that our rate of unexpected unknown-cause infant deaths has dropped.


[Comments:]

2006-11-14:

Australia /Emma King A/:
The Queensland Child Death Review Team has also experienced this problem. Sudden Unexpected Death in Infancy is a research classification and does not correspond to any single medical definition, International Classification of Diseases categorisation or Australian Bureau of Statistic's categorisation.

Rather, the aim of this grouping is to report on the deaths of that group of apparently normal infants who die suddenly for reasons often unknown - particularly at the time of death. The SUDI grouping therefore includes deaths associated with infections or anatomical or developmental abnormalities not recognised before death, sleep accidents due to unsafe sleep environments, and deaths which initially present as sudden and unexpected but are revealed by investigations to be the result of non- accidental injury. It is also deaths due to SIDS and where a cause of death is not determined. Cases of SUDI are usually explained at post-mortem and those that are left unexplained are classified as 'undetermined'.

In March 2004, SIDS and Kids Australia (the peak non-government body on SIDS) hosted the first National SIDS Pathology Workshop. The workshop brought together paediatric and forensic pathologies from each state and territory with the aim of obtaining a national consensus for the common definition of SIDS and a recommended standard autopsy protocol for infant deaths. Participants at this workshop agreed to implement the adoption of the 'SIDS Redefinition Conference' held in San Diego in January 2004, at which the following new definition of SIDS was developed:
'The sudden, unexpected death of an infant under one year of age, with onset of the fatal episode apparently occurring during sleep, that remains unexplained after a thorough investigation including performance of a complete autopsy and review of the circumstances of death and the clinical history' (Krouse et al. 2004:235).

It was also agreed that stratifying cases of SIDS into subcategories would provide diagnostic guidelines, identify and include cases that had previously been excluded because of shared sleeping or stomach-down sleep position, and separate cases based on the degree of certainty with which a diagnosis of SIDS can be made (Krous et al. 2004:236-37). Thus the San Diego definition includes the following subcategories:

* Category IA SIDS: The classic features of SIDS are present and completely documented, including the typical age range of 21 days to less than 9 months. Cases in this category show no evidence of trauma, disease or lesions at autopsy and all toxicology results are negative. Normal clinical history growth and development and the death scene show nothing that could have contributed to the death.

* Category IB SIDS: As above, except an investigation of death scenes was not performed and at least one the following analyses was not performed:
toxicologic, microbiologic, radiologic, vitreous chemistry, or metabolic screening studies.

* Category II SIDS: Meets the criteria for IA or B, except the infant was outside the typical age range; had abnormal growth or development; an autopsy reveals an abnormality that appears significant, but whose precise role in the death is difficult to determine; or mechanical asphyxia due to overlay cannot be excluded.

In addition, the San Diego forum proposed that those cases that do not fall into the above categories, but in which "alternative diagnoses of natural or unnatural conditions are equivocal (including cases for which autopsies are not performed)", be attributed to 'unclassified sudden infant death' (USID) (formerly undetermined or unascertained) (Krous et al. 2004:236).

Cases of SUDI should therefore be classified as undetermined if:
* natural disease processes were detected that are not considered sufficient to cause death but that preclude a diagnosis of SIDS
* there are signs of significant stress
* non-accidental but non-lethal injuries were present, or
* toxicologic screening detects non-prescribed but non-lethal drugs
(Mitchell et al. 2000:311).

Participants in the March 2004 National SIDS Pathology Workshop also agreed that the ill defined terms 'unascertained' and 'undetermined' would be replaced by 'unclassified sudden infant death (USID)'. The Commission is concerned that the introduction of the acronym USID and the similarities between this and the terms ' sudden unexpected death in infancy (SUDI)' and ' sudden infant death syndrome (SIDS)' may cause unnecessary confusion among pathologists, coroners, coders and health professionals. To demonstrate (similar cases to those identified in New Zealand) pathologists in Queensland have certified the cause of death of infant deaths as 'sudden unexpected infant death syndrome (2 deaths, 1 of which was an infant over 12 months old)', 'unclassified sudden infant death syndrome (1 death)', and 'unclassified sudden infant death (1 death)'. The Commission will write back to each pathologist/ coroner and to Queensland Health and SIDS and Kids Australia to resolve these ambiguities. In the interim the Commission has decided to continue to use the term 'undetermined' to refer to those infant deaths where following post mortem examination a cause of death was unable to be ascertained. For the above cases the Commission has coded:

- the 'sudden unexpected infant death syndrome (2 deaths, 1 of which was an infant over 12 months old)' as R99 as the definition of SIDs precludes its application to infants over 1 year and R95 as the infant's autopsy was unremarkable and the Commission considered this to be the intention of the pathologist when certifying the death,

- the 'unclassified sudden infant death syndrome (1 death)' as' other and ill-defined causes of mortality' R99 because of the presence of the word 'unclassified'

- and 'unclassified sudden infant death (1 death)' as R99 in line with the outcomes of the First Pathology Workshop. All of these cases are however being queried with the pathologist and Queensland Health to confirm.

The Commission would encourage others to also write to pathologists, coroners and key agencies to query causes of death so as to ensure as Christine in New Zealand has identified, that we do not artificially lower the rate of SIDS deaths and/or give a false impression that our rate of unexpected unknown-cause infant deaths has dropped.

Australia /Jeanine Young/:
I was speaking to Peter Fleming and Peter Sidebotham about this issue when visiting their departments in the UK in Sept. They issue a word of warning - the pathologists' definition of SIDS from the San Diego Conference may have been published however it has not been accepted by all countries in which infant death reviews are conducted by a range of professionals, including, but not exclusive to pathologists. In particular, the UK and Europe are not using this definition for classifying SIDS deaths. For example the present Qld system for stating SIDS (cosleeping) on a birth certificate is viewed as rubbish by many SIDS experts given that cosleeping may have been a risk factor but not a cause of death. To state SIDS indicates insufficient evidence for cause of death but by stating cosleeping implicates that in the SIDS death. To state cosleeping - would it have been better to classify as a sleeping accident if evidence of mechanical suffocation; therefore not a SIDS? There is no category for SIDS (maternal smoking) or SIDS (paternal alcohol or substance abuse).

In particular these countries have trouble with the new age limit of 12 months - so continue to use the 1994 Stavanger definition.

Thoughts to stimulate discussion only

Brazil /Ruy Laurenti/:
In Brazil both terms are used by pediatricians and medico-legal doctors. In the death certificates both terms are declared. We code R95 in all the cases.

Canada /Patricia Wood/:
Provided the decedent is under one year of age at death, we would code SUDI to R95.

FYI - Although there are no inclusion terms for ICD-10 code R95 in Volume I, there is a list of inclusion terms, including SUDI, in NCHS Instruction Manual 2b, Section IV, Part G:

Sudden infant death syndrome (R95)
Includes:
Cot death
Crib death
SDII, SID, SIDS, SUD, SUDI, SUID

Sudden (unexpected) (unattended) (unexplained)
death (cause unknown) (in infancy) (syndrome)
infant death (syndrome)

Causing death at ages under 1 year

Excludes: The listed conditions causing death at ages one year or over (R960)

2006-10-23:

Germany /Orlando Özer/:
We would code SUDI R95. It does not seem appropriate to code it R98 or R99. One should resolve if the term SUDI is really used as an exact synonym for SIDS by the coroners or pathologists in the respective country.

Norway /Gunvor Ostevold/:
This is a very interesting topic. In Norway we do not have the term SUDI, but forensic autopsies often use the term Borderline SIDS. It may be similar to SUDI. It means there is no clear evidence for any cause of death, but the autopsies do find some vague symptoms like nasopharyngitis or "something" in the nervous system, but this symptoms can not cause the death. We use R95 for both terms, SIDS and Borderline SIDS. I am looking forward to the further discussion!

2006-10-24_01 Violent deaths registered before external cause info is avialable

(Chapter XX)

[Question from Elaine Tower, England:]

In England and Wales we have deaths which are registered before final details of the cause are known - these are a special case for deaths where the inquest (investigation) into the death is delayed whilst awaiting the outcome of police investigations into third-party involvement in the death.

The majority of these eventually turn out to be homicides (over 90%.) Some eventually become land transport accidents or rarely other types of accident. At present we are assigning Y339 as the Underlying cause. This is largely historical and corresponds to the code we used in ICD-9, E988.8. This system was set up in 1979, so we did not have this type of death before then.

Unfortunately because of this these deaths are then being classed together with poisoning / injury deaths undetermined intent. Although in our national statistics we can deal with this and draw people's attention to it, we cannot do this when others use the data without any guidance from us.

When the final outcome is known we then receive further information which describes the outcome ie Homicide; Killed unlawfully; Dangerous Driving etc.

For us to differentiate these deaths before the final outcome is known we are considering using a national "U" code until the final outcome is known and then we will re-code accordingly.

We have the following questions:
1) Do other countries use a system like this?
2) If so, what codes are they using?
3) We are thinking of using perhaps U559 code (this has not been agreed yet here by us so I am using it as an example). Is there a route we need to follow to use this particular U code? Or can we use it just for our own national statistical purposes?


[Comments:]

2006-11-14:

Brazil /Ruy Laurenti/:
In Brazil the codification of the u.c. of death is decentralized and is made at the municipal level. In the majority of the municipalities the coder waits (approximately one month ) to receive the confirmation of the type of the external cause: homicide, transport accident, etc and then code the u.c.of death. I do not understand why use U55.9.

Canada /Patricia Wood/:
We have a system very much like what Elaine is describing. Each Canadian province and territory has a coroner or medical examiner system which is legislated to investigate sudden and unexpected deaths. Often these events are registered before the investigation is complete. These certificates are usually labeled as "pending" or "interim" and we have traditionally coded them to R99 (or 799.9 in ICD-9). We do update the national mortality file as the final results are made available to us, but sometimes it is so long after the event that we have published the statistics including the "pending" or "interim" cases in R99. We know that this overstates our deaths due to unknown cause and under reports the actual causes, but as Elaine points out, data users are not always aware of this fact. I think that using a U code to identify such cases is an interesting idea.

WHO /Robert Jakob/:
I think we should suggest to mark such cases in a different fashion. To add a national code for "on hold" is not the corret way.

2006-11-21_01 IHD due to COPD

(Rules GP)

[Question from Christine Fowler, New Zealand:]

We often find doctors certifying ischaemic heart due to Chronic Obstructive Pulmonary Disease, for example:
1(a)        Acute myocardial infarction - minutes
1(b)        Ischaemic heart disease - 3 years
1(c)        Chronic Obstructive Pulmonary Disease - 15 years  

The ACME decision tables do not accept COPD as a cause of IHD and therefore, in the above example, commence the sequence at 1(b) and select the M.I. as the underlying cause. Do we just accept that this is the correct way to code this scenario, or should we accept that the sequence in which the certifying doctor has entered the causes of death is legitimate (in the face of new medical knowledge) and code the COPD as the underlying cause?  

Are other countries finding an increase in the certification of COPD as the underlying cause of IHD?


[Comments:]

2006-12-05:

Norway /Gunvor Ostevold/:
Yes, we have seen an increase of sequences like this. Earlier, when we did the classification manually we did not agree with this sequence and coded
1a) I219 as the underlying cause. We did accept heart failure due to chronic obstructive pulmonary disease but not ischaemic heart disease. On the contrary I have seen ACME accepts COPD due to heart failure and this is an sequence I do not understand.

Sweden /Lars Age Johansson/:
I have seen such certificates too, although I don't know if they are more common now than before. Perhaps this is something that the Mortality Reference Group should have a look at. Apparently there is some epidemiological support for the idea that COPD causes changes to arteries in many parts of the body. Any cardiologist or pulmonologist on the Mortality Forum who knows more about this?

2006-11-21_02 Dehydration an obvious consequence of dementia

(Rules R3)

[Questions from Gunvor Ostevold, Norway:] 

How do you deal with a certificate like this:
1a Dehydration, E86
II Dementia, F03

ACME selects dehydration as the underlying cause of death, but we would like to use rule 3 in this case and select dementia as the underlying cause of death. How do you code this? 


[Comments:]

2006-12-05:

Canada /Patricia Wood/:
We would select dehydration as the underlying cause of death. Rule 3 states that dehydration may be assumed to be a consequence of any intestinal infectious disease, but does not suggest that dehydration is obviously a direct consequence of dementia. If dementia had been certified in Part I as the cause of dehydration, we would apply the General Principle to select dementia, but not when reported in Part II.

2006-11-28:

Sweden /Lars Age Johansson/:
This is an interesting question! Obviously there is a link between the dehydration and the dementia. The question is if dehydration is an obvious consequence of dementia. I would say that it shouldn't be - if a demented person is properly cared for, he or she shouldn't die from dehydration. So if we want to focus on the quality of the care, I would say that dehydration is a more "useful" underlying cause than dementia.

2006-11-21_03 Tumor cerebri NOS - malignant or not?

(Chapter II)

[Questions from Gunvor Ostevold, Norway:]

Very often physicians use the term "Tumor" whether it is benign or malign. The alphabetic index points it out as a tumor with morphology with uncertain behavior. Do you code D432 for "Tumor cerebri"? Or is it more likely it is malignity and the code is C719?


[Comments:]

2006-12-05:

Canada /Patricia Wood/:
In the absence of an implication of malignancy, we code the lead term tumour as indexed, so we would code tumour cerebri as D432. If you feel that some or all of the "tumours" reported are malignant, are you able to confirm that with the certifier(s)? That might serve two purposes; it could improve the quality of your data and provide your certifier(s) with some instruction.

Norway /Gunvor Ostevold/:
Thank you for your response of my question. An additional thought after reading Mr.Hasiak's interesting answer: I'm impressed of his well organised Mortality Forum register! How can you remember a question discussed in 1997?

To my excuse I was not working with causes of death at that time! However, I have more than one time asked about subjects which have been discussed earlier. So, my question is: is there a web site for Mortality Forum where I can find former questions and answers organised after topics? Before asking the Forum, I could look up this web site.

Sweden /Lars Age Johansson/:
An archive of all previous discussions is available at the Nordic Classification Centre's website, at http://www.nordclass.uu.se/index_e.htm. Click "Mortality Forum" at the left, and then "Index to the Mortality Forum". There is plenty of reading there - we have discussed over 600 issues by now!

2006-11-28

Brazil /Augusto Hasiak Santo/:
Regarding the question from Gunvor Ostevold, as we see above, I would like to remember that a similar one was presented by Mary Heanue and discussed on 2 -IV-1997.

At that time I commented that: "Brazil /Augusto Hasiak Santo/:
Regarding the question from Mary Heanue, Ireland, about assuming tumors as malignant, I would like to inform that in Brazil, since the 8th Revision, the words "tumor" and "neoplasm", are taken as "malignant tumor" and as "malignant neoplasm" when appearing in death certificates. This decision was done after the results of the Interamerican Studies of Mortality (Adults in 1962-1964 and Childhood in 1968-1969), in which the death certificates were revised with data from the families of the dead, hospital records, autopsies data and all other available information, providing a new and "more correct" underlying cause of death. It was found that when the physicians mentioned tumor or neoplasm, they really wanted to mean malignant tumor or neoplasm.

It was noticed that when a benign neoplasm caused the death, this fact was so unexpected that the physicians mentioned them as benign or informed the morphological type, such as "adenoma" or "meningioma". After those two investigation, many other studies have given the same results, with the same methodology, that we use to call "Puffer Method", after Ruth Rice Puffer, who coordinated the Interamerican Studies." We continue to assume the same interpretation for the term "tumour".

Sweden /Lars Age Johansson/:
Augusto's reply shows that it is extremely important to investigate how ambiguous terms are in fact used by the certifiers. We shouldn't take for granted that physicians always use them in the same sense as the English Alphabetical index of ICD. However, results from one language or area might not apply to another part of the world. In Sweden we usually query the unspecified tumours, because "tumour" here quite often refers to a non-malignant tumour. A (perhaps not very useful) piece of advice would be to query these cases till you have got a fair idea of how the expression is used in Norway. Then you can make a decision whether to code them as malignant, or to continue querying them.

2006-11-21_04 When to implement a new ACME version

[Questions from Gunvor Ostevold, Norway:]

We are now using Acme version 2006.2 coding 2005 data. I'm wondering when other countries implement new versions. Do you implement new versions when they are released or do you wait until you start coding a new data-year?


[Comments:]

2006-12-05:

Canada /Patricia Wood/:
We debated this issue shortly after we implemented ICD-10 and our decision was to use the annual MMDS version that matched the data year we were coding (i.e. MMDS Version 2006 for 2006 deaths). We implement any new versions (.02, .03, etc.) of that annual version when we receive them, but we do not use the newest annual version until we are coding deaths from that year.

Our reasoning for this is that we will be better able to explain any variations in our trends that are caused by changes to the interpretation or application of the Classification and/or the rules for selecting underlying cause of death.

2006-11-28:

Sweden /Lars Age Johansson/:
In Sweden, we don't change ACME version during the data year. Shifts in trends and levels are much easier to keep track of if you apply the same classification procedures to the entire data year. When we start working on a new data year, however, we will implement the most recent version of ACME, even if this might be somewhat ahead of our new data year. For example, we are now using the 2006.2 version to process our 2004 data. The drawback is that we sometimes will start using a new ICD code before its official implementation date, but the advantage is that we can benefit from all other corrections and improvements to ACME much sooner.

2006-11-21_05 Code for sequelae of brain injuries due to boxing

(Chapter V, XX)

[Question from Lars Age Johansson, Sweden:] 

We recently had a certificate something like this:
1(a)    Pneumonia
1(b)    Dementia
1(c)    Sequelae of repeated brain injuries 

According to additional information on the death certificate, the repeated brain injuries were due to the fact that the patient had once been a semi-professional boxer. How do you code injuries someone has brought down on himself, more or less voluntarily? 


[Comments:]

2006-12-05:

Canada /Patricia Wood/:
Although the decedent stepped into the boxing ring voluntarily (or so we would assume), his intention was likely to avoid injury! I suggest we code the external cause of the brain injuries as accidental, and since they were sustained in the past, as sequelae of an accident ... Y86.

Norway /Gunvor Ostevold/:
I have never seen a sequence like this, so it must be an unusual cause of death. It is not intentional self-harm nor is it an assault. My suggestion is therefore W50, but this code does not cover exact what happened.

2006-11-28_01 Septic shock following emergency caesarean section

(Chapter XV)

[Questions from Elaine Tower, England:]

1a Septic shock due to methicillin resistant staphylococcus aureus
1b Uterine sepsis following emergency caesarean section delivery at term.
We have coded to O988; A410; O861 with an underlying cause of O861. Is this what other coders would have chosen?


[Comments:]

2006-12-05:

Canada /Patricia Wood/:
Elaine, I think that maternal deaths are tough to code ... since the uterine sepsis is certified as "following" the caesarean, I wonder if O85 Puerperal sepsis might be a good code for the underlying cause of death, with B95.6 to capture the Staphylococcus aureus as the infectious agent?

2006-11-28_02 Cepacia syndrome and cystic fibrosis

(Rule R3)

[Questions from Elaine Tower, England:]

1a Cepacia syndrome
1b Pneumonia
II Bilateral Lung Transplant for Cystic Fibrosis
We have coded A498; J958; Y830; E840 with an underlying cause of E840

Should we be coding the Cepacia syndrome as the reason for the Transplant or should we think the Cepacia syndrome was developed after the Op? According to the Cystic Fibrosis information from the web it states "It occurs in a minority of patients with Cystic Fibrosis"


[Comments:]

2006-12-05:

Canada /Patricia Wood/:
I would certainly code the underlying cause as E84.0 Cystic fibrosis with pulmonary manifestations, considering the pneumonia and Burkholderia cepacia to be complications of the lung transplant and/or the cystic fibrosis.

Germany /Orlando Özer/:
We'd also code E84.0 as Underlying Cause.

2006-11-28_03 Code for congenital cardiomyopathy

(Chapter IX, XVI, XVII)

[Question from Jan Kardaun, The Netherlands]:

Case: Neonate, died within days after birth; preterm with birth weight corresponding to duration.
1a) circulatory insufficiency, 24 hr
1b) (probable) congenital cardiomyopathy COD certificate completed by a neonatologist.

Options (in the order of our preference):

Q24.9 [Congenital malformation of heart, unspecified / Other congenital ...diseases NOS of heart]:
It is in the group of malformations of heart, even though this is not a malformation in the strict sense/

I42.4 [Endocardial fibroelastosis / Congetinal myocardiopathy]:
Is the most specific for the mentioned disease (congenital cardiomyopathy) There is no age restriction of I42

P29.0 [Neonatal cardiac failure]:
Expresses best that it is a neonate, but excludes Q20-Q28

See also Vol II 4.2.9: heart disease NOS under 4 weeks may be assumed to be congenital (but give no coding)


[Comments:]

2006-12-05:

Canada /Patricia Wood/:
I would code the underlying cause of death to I42.4 Endocardial fibroelastosis, which includes congenital cardiomyopathy. Also important is the exclusion note at Q24; excludes: endocardial fibroelastosis (I42.4)

Germany /Orlando Özer/:
I42.4 is the adequate code for 'congenital cardiomyopathy' according to the Index. We'd choose it as UC.

2006-12-05_01 Code for Berdon's syndrome

(Chapter XVII)

[Question from Elaine Tower, England:]

We have the following death which we are interested to see how other countries would have coded:
Female baby 6 weeks old
1a Multi organ failure
1b Megacystitis, Microcolon Intestinal Hypoperistalsis syndrome
Codes used are R68.8; Q43.8; R19.2 with underlying cause of Q43.8

We have only coded the Microcolon. Would other countries have also used a Bladder Anomaly code? There is some understanding that it is also known as Berdon's syndrome - As this is a very rare condition that is not listed in ICD10 updates, should it be considered for the future?

Thank you.


[Comments:]

2006-12-20:

WHO /Robert Jakob/:
This rare syndrome was reported in less than 25 patients in the world literature, and in 9 of the papers on the subject two affected sibs were reported. The disease is thought to be probably transmitted autosomal recessive and relevant genes have been identified on the chromosome 15q [orphanet]. The lack of certainty regarding genetics and the dominant intestinal manifestations would justify to code to Q43.8.

The statistical impact of such cases will be extremely low.

2006-12-12:

Canada /Patricia Wood/:
Of interest, megacystitis is indexed in NCHS Instruction Manual Part 2e, Volume 3 (2006) to ICD-10 code N328, prompting me to run this case through MMDS Version 2006.02. MMDS assigned megacystitis to N328 but did not code intestinal hypoperistalsis syndrome (U00) and generated a MICAR reject. I resolved the reject by coding intestinal hypoperistalsis syndrome (which is not indexed) to K598. ACME then proceeded to apply Rule 1 and select N328 as the underlying cause of death. It did the same thing when I coded intestinal hypoperistalsis syndrome to R192.

I suspect that this is not actually appropriate given what Elaine tells us about the reported causes of death being symptoms of Berdon's syndrome.

This presents us with a recurring problem - how should we code syndromes that include congenital anomalies of more than one system or site? One possibility is Q878, Other specified congenital malformation syndromes, not elsewhere classified, but there is a very real danger that this would end up being the code for a very long list of assorted congenital anomaly syndromes.

 

 

Comments