Old Topics‎ > ‎


2004-01-13_01 Fournier's syndrome obvious consequence of tumour of rectum

(Rule 3)

[Question from Augusto Hasiak Santo, Brazil:]

This death certificate includes Fournier syndrome as a cause of death in a woman of 70 years of age. Fournier syndrome in women has been already discussed twice in the Forum (1998 -05-25 Q3 and 2001-09-10 Q1). In the last discussion a proposal of forwarding the question to the WHO Update Reference Committee was done. This death certificate introduces another question: is Fournier syndrome a condition that can be considered as a direct consequence of neoplasms aiming at the application of Rule 3? Please underlying, non-underlying causes and how to code Fournier syndrome in women.

Female, 70 years of age

I (a) Multiple organ failure
I (b) Septicaemia
I (c) Fournier syndrome
I (c)
II     Advanced tumor of rectum with metastasis
[Suggested coding:]
  Underlying Multiple causes
Sweden C20   R688/A419/N768*C20   

Canada /Patricia Wood/:
I agree with Lars Age's codes for the conditions on this certificate, including the soon-to-be implemented code for Fournier's syndrome in females, but I am inclined to leave the underlying cause of death as N768. I hesitate to apply Rule 3 to reselect the underlying cause of death as C20 because I am do not know if Fornier's syndrome is an obvious consequence of a malignant neoplasm of the rectum.  If the relationship between Fournier's syndrome and malignant neoplasm of the rectum is such that Rule 3 can be applied, the ACME decision tables do not reflect it.


Sweden /Lars Age Johansson:/
The Mortality Reference Group did arrive at a suggestion for Fournier's syndrome in women last year, and the suggestion was accepted by the Update Reference Committee at its meeting in October. The code will be N76.8, which we believe is the best parallel to the code N49.8 for males. The official implementation date is January 2005.
In this case I would apply Rule 3, since the site of the gangrene is so close to the site of the primary tumour.

2004-01-13_02 Metastatic ovarian carcinoma with carcinoma of breast

(Rule 4,2,7)

[Question from David Lidsky, Israel:]

1a metastatic ovarian carcinoma
1b carcinoma of breast

My colleague Mariana Yardeni coded the above certificate: C79.6/C50.9, underlying cause C 50.9. The MMDS software agreed with her.
My coding was: C56/C50.9, underlying cause C97.

Example 21 on page 78, Volume 2, seems clearly to imply that "metastatic ovarian carcinoma" should be regarded as referring to a primary ovarian tumor, and coded C56. The question is whether this initial interpretation, and coding, should be modified (with the consequent change in underlying cause) as the result of a primary malignancy on the next line. Is there an instruction in ICD-10 telling us to make such a change?

More generally, to what extent should our understanding  of a specified diagnosis be influenced by what appears elsewhere on the certificate? That is a question which we ask ourselves all the time.

Canada /Patricia Wood/:
Although the instruction and example that David cites suggest that metastatic ovarian cancer should be coded to C56, I believe that it should be applied when there is only one malignant neoplasm reported on the certificate.   The instruction and example that Lars Age refers to seem to be applicable to this particular certificate, suggesting that C509 should be selected as the underlying cause of death.  However, I totally support Lars Age's idea of requesting further clarification from the certifier.  Although I believe the "correct" application of the neoplasm instructions leads to C509, I am not altogether sure that this is the true underlying cause of death.  One of the new things in ICD-10 was that we were no longer going to take account of order of entry for malignant neoplasms, but in this case that is exactly what selecting C509 would imply - that a primary malignant neoplasm of the breast caused ovarian metastases.  Hmmm.


Sweden /Lars Age Johansson/:
Instruction (iv) on page 79 could also be applied to this case. It says that "if more than one site ... is mentioned and all but one are qualified as metastatic ... code to the site that is not qualified as metastatic". Following that instruction, we would get breast cancer as the underlying cause of death. So I think that both Ms Yardeni and the MMDS have followed the coding instructions here. Is there any possibility to go back to the certifier and check if the patient had two primary tumours? Perhaps that would be the best thing do.

It is difficult to give a straightforward answer to David's question on information appearing elsewhere on the certificate. In Sweden, we generally use information appearing elsewhere on the death certificate to determine etiology and duration of a condition. For the malignant neoplasms, however, we use the instructions in Vol 2 to decide if a tumour is to be coded as primary or secondary, not the order of entry.

2004-01-13_03 "Immune compromised" meaning HIV?

(UC Chapter I)

[Question from Trinidad and Tobago to the Forum-CIE, forwarded by Roberto Becker, PAHO:]

In the Caribbean doctors are reluctant to report HIV/AIDS on the death certificate, the result being under-reporting of HIV/AIDS deaths. However, we see terms like 'immune compromised' on the death certificate which may or may not be reported along with an HIV/AIDS opportunistic disease. Would it be valid to assume that the underlying cause of death is HIV/AIDS, especially if there are opportunistic diseases as well as terms like 'immune compromised' recorded on the death certificate, but HIV/AIDS is not on the certificate? This is distinct from Rule 3 which states that HIV must be reported on the death certificate.


PAHO /Margaret Hazlewood/:
The question from Trinidad and Tobago and the ensuing discussions based on the ICD-10 framework are very insightful as a basis for developing consensus in identifying correctly HIV/AIDS deaths. Under-reporting of HIV/AIDS cases in the Caribbean is "real" because certifying physicians, for many reasons, will NOT state HIV/AIDS as the underlying cause of death even when it is well-known in the community and from the clinical records that the person's demise was associated with the HIV/AIDS status. Selecting, coding, and reporting the underlying cause of death in these cases are problematic and can have legal implications for the medical coder. Coding is done from the source document (the death certificate; however, quite often the deceased records are available for verification. Quite often, querying the certifier will not make a difference since usually the death certificate is never amended. The physicians are fully aware of the importance of accurate reporting of HIV/AIDS deaths; protection of the family and confidentiality issues appear much more important. The stigma attached to HIV/AIDS remains a major concern/challenge in the Caribbean.

[From our Spanish-speaking twin, the Forum-CIE, we have received the following comments. Many thanks to Roberto Becker for translating and forwarding these very interesting comments! /Lars Age]

Nicaragua /Velia Emperatriz/:
In the case of Nicaragua, the problem is the same, the doctors use terms like "immunodepressed" some times together with opportunistic diseases, but mostly just that kind of term. We continuously try to fix this problem in workshops with the physicians, but in fact we did not reach a consensus yet. We should not "hide" the data for "ethical" reasons.
Maybe WHO could address this problem and propose a more technical solution.
Chile /(Danuta Rajs/:
In Chile it is exactly the same. The doctors, for different reasons, including request of the families, use to register things like "T - lymphocyte deficiency" or "immune disease". Our procedure (not exactly from the rules) is:
a) the coders have a list with terms that can be related to AIDS
b) we query CONASIDA (National Committee for AIDS), including virus lab, to check if the case was reported. If not, we code B20-B24, according to the complications mentioned in the death certificate
c) the cases not reported are lost in the first moment, but some of the information we can get later
d) in final data validation we use to discover some more cases, among sepsis, miliary tuberculosis and other things.
I suppose that some transient rule should be suggested, but no all the countries have this type of coordination, available in Chile. In my opinion, it is better to code in B20-B24, in spite AIDS was not specifically mentioned.

Bolivia /Herland Tejerina/:
I think the key point is to assign a conventional code (D89.9?) to the cause reported and let to the one analyzing the data the decision to considere those cases (all or a %) as AIDS. I think it's dangerous to assume always HIV when there is no evidence in the clinical records.
Mexico /Luis Manuel/CEMECE/:
My opinion is that only those cases where the epidemiological investigation allowed to confirm this cause (recorded in the certificate) should be coded as HIV/AIDS. This decision should not be let to the coder.
If this is the practice in the country it is advisable that all this cases be epidemiologically investigated and that the physician doing the investigation be the one defining if it was HIV/AIDS and report to the coder to amend the underlying cause.

Cuba /Ana Consuelo/CECUCE/:
This is a very serious problem and any solution gave from outside the specific country may complicate the problem. However I will give my criteria:
1. Why are them reluctant, is there any legal or social reason to act like that? The first step would be to solve the origin of the problem, that is to report the disease according to the case.
2. Not all immunodeficiency followed by opportunistic diseases are AIDS cases, in spite there is a high probability.
3. It is also important to consider the age of the deceased. The probability is higher in young people.
4. How much are the country gaining or loosing. I mean, how many are being coded as AIDS in excess and how many less, depending on the criteria adopted?
5. In ICD instructions there is nothing supporting what is being proposed. However, I believe that in some circumstances one country can decide. For example, since ICD-9 we use to code septic shock in 038.9.
I dare to anticipate Prof. Becker's answer (or part): "Code what is informed in the death certificate and only when analyzing the data redistribute the cases, considering those elements..."
Is that the best? The final decision is up to the country.

Honduras /Guillermo Guibovich/:
The same occurs in Honduras, the reluctance to report HIV/AIDS as cause of death. I think that it would be valid to assume HIV/AIDS in the mentioned cases.

Spain (Galicia)/Sara Cerdeira:
My opinion is that we should not assume in any case the cause of death; neither in this case, HIV/AIDS in the death certificates with 'immune compromised', still associated with opportunistic disease (because it is what is expected) regardless the origin of the immunodeficiency.

Brazil /Ruy Laurenti/:
The Brazilian Collaborating Center agrees to consider AIDS as the underlying cause in the cases mentioned by Trinidad & Tobago.

Argentina /Lidia Turina/CNCE/:
Te Death Certificates containing deficient information like that, need to be amended to avoid the lost of information about AIDS and assumptions on the selection of underlying cause of death. We have to insist with the doctors that it is up to them to obtain better information.
Immunodeficiency syndromes, congenital or acquired, very often occur together with infections, autoimmune diseases and lynphorreticular neoplasms. To assume the existence of HIV/AIDS is equivalent to confirm the presence of the disease, mistake that the coders should not make. Rule 3 does not apply and if only immunodeficiency is informed, the cause should be coded to D89.9.

Cuba /Miguel Ángel/CECUCE/:
I don't think it's correct to assume the existence of HIV/AIDS when the physician does not certify. There are multiple causes that can lead to immunodeficiency and then to an opportunistic disease. It's important that, based on specific investigations, the users be informed about the possibility of under-reporting, and the health managers about the reasons of under-reporting, in order to fix the problem.

USA /Carol Lewis/:
In some of the countries in which I've worked, death certificates are used for purposes that have little to do with public health. For example, radio stations may require that a death certificate be presented before the death and funeral arrangements can be announced. It is therefore understandable that physicians concerned about patient (and family) privacy are unwilling to include diagnoses such as HIV/AIDS, alcoholism, suicide on the certificate but will provide this information confidentially in response to a direct request for clarification.
I believe that in cases such as the one cited in the questions, the correct action to take is to query the doctor who signed the certificate. I am concerned that once coding decisions are made based on assumptions for this condition that the practice could be extended to other situations as well.

PAHO /Rafael Mazín - Regional Advisor on HIV/AIDS Prevention and Comprehensive Care/:
We should not use as a definite fact a diagnosis assumption because immunodeficiency may be acquired or associated to underlying conditions not necessarily HIV/AIDS.

Roberto Becker/PAHO:
I agree with the majority of the opinions. I do not recommend to code as HIV/AIDS the cases mentioned above. The reasons are very well explained in several answers. Ana Consuelo's "anticipation" is correct... The same is mentioned by Herland. I think that for the cases mentioned in the Question D89.9 is OK. It's also important to note the opinion of Rafael Mazín, who is exactly managing AIDS in PAHO.
In various discussions, mainly in Forum-CIE, I said that when tabulating and analyzing data, not when coding, we may suppose and interpret as we want.
The problem related to "sensitive" or "avoided" diagnoses is real and occurs for several reasons. Carol mentions one example. I don't think we can have a general solution for all cases (AIDS, suicide, alcoholism, etc.) and also the solution will not be the same in every country.
Of course looking for complete information querying the certifier is always the best way to get better information.
In the case of AIDS, as mentioned by Danuta, it is very important to maintain a good coordination with the specific program to, starting with key words, verify if the case was known, and get the confirmation. The same is valid for "flaccid paralysis" (poliomyelitis), "jaundice" (yellow fever or hepatitis), and many others.
However, for many diseases or conditions there will not be a "program" to ask, nor key words to start with.
In some areas, maybe the solution could be some kind of agreement so the certifiers can report the "sensitive" diagnoses in a confidential letter. It also could be agreed to report the diagnoses (in the death certificates) with key words previously agreed, but not with ICD codes, because the coders, not the certifiers, are supposed to code.
In summary, a specific procedure has to be established for each area and each case.


Sweden /Lars Age Johansson/:
We had the same problem in Sweden when the HIV epidemic started, but fortunately the matter is far less sensitive now and the problem has more or less disappeared. However, I would be reluctant to code HIV if the certificate only mentions "immune compromised", since you can develop an immunodeficiency for many other reasons than HIV (malignant neoplasms, for example). I don't think the presence of an opportunistic disease makes a very great difference - wouldn't an individual with a severe immunity deficiency risk to develop an opportunistic infection, regardless of the cause of the immunodeficiency? And then we have the ICD instruction that tells us not to code assumed intervening causes (Vol 2, p 66). If we are not allowed to code assumed intervening causes, then I cannot see how we could code an assumed underlying cause.

Still, some of the 'immune compromised' deaths are, in all probability, due to HIV. In Sweden, when the certifier had used 'immune compromised' or similar expressions, we always queried the certificate and asked for a more specific description of the condition. Would that be an option in Trinidad and Tobago as well?

2004-01-27_01 List of non-specific codes

[Question from Sue Walker, Australia:]

I am looking for a list of codes in ICD-10 that can be considered non-specific - its not as easy as ICD-9 when we could say that all .8s and .9s were non specific. Does anyone have a list that they could send me?



Canada /Patricia Wood/:
Sorry, Sue.  We (Statistics Canada) does not have such a list and I checked with Lori Moskal at CIHI and she said they don't have one either.  Maybe this is a job for an enterprising student of the Health Information Management program?

2004-01-27_02 Code for HIV positive

(UC Chapter I)

[Question from Ruy Laurenti, Sao Paulo Classification Centre:]

We are submitting to the Mortality Forum the following question: how should we code the information "HIV +" in death certificate of newborn and in certificate of stillbirth?

Case 1: Stillbirth
I - a) - Eclampsia (mother)
b) - HIV+
c) -
d) -
II - -

Case 2 : Male, 7 days
I - a) Failure of multiple organs
b) Pulmonary haemorrhages
c) Septicemia
d) -
II - Prematurity
HIV positive

Case 3: Female, 3 days
I - a) -
b) Sepsis
c) Prematurity
d) Premature rupture of membranes
II - Mother HIV +

Canada /Patricia Wood/:
I, like Lars Age, would select P000 as the underlying cause of stillbirth on the first of Dr. Laurenti's cases and I would select P011 as the underlying cause of death for the third one.

I am not sure who is HIV+ in the second case - the mother or the baby? Either way, I would select P369 as the underlying cause of death. 

When reported on a death certificate, we code HIV+ to R75, so we can not code it to P00.2 as an infectious or parasitic maternal condition affecting the fetus or newborn when reported on a stillbirth or infant death certificate.  HIV+ is not really a "condition" per se, but I suppose that when reported about the mother on a stillbirth or infant death certificate, we could code it to P00.8?  That is not much of a "flag" though since it an unspecified code, but I think that to code R75 on a stillbirth or infant death certificate suggests that the baby was HIV+ rather than the mother.

We do not use the Z codes on our mortality database.


Sao Paulo Classification Centre /Ruy Laurenti/:
Regarding the question that the Brazilian Center sent to the Mortality Forum. We agree and did the same as Lars. However we do not understand why he put the code O98 that is for maternal causes or mortality and not to be used for newborn death or stillbirth.

Sweden /Lars Age Johansson/:
Sorry, the paragraph on O98 was a word processing mistake! I have now corrected


Sweden /Lars Age Johansson/:
The certificates do not mention HIV disease, only that the women (and in the second case perhaps the fetus as well) were HIV positive. That being the case, I am reluctant to use a code in B20-B24. On the other hand, recent studies suggest that HIV positive women on antiviral therapy run a far greater risk to develop eclampsia or to give birth prematurely. Quite obviously it is important to record in some way that the woman was HIV positive.

In the first case I would select, as the underlying cause code, the usual code for fetus affected by eclampsia in the mother (P00.0). At first P00.2, fetus and newborn affected by maternal infectious and parasitic diseases, might look like an alternative. However, neither Z21 (which includes HIV positive) nor R75 (laboratory evidence of HIV) count as infectious diseases. My suggestion is to use P00.0 for the underlying cause, and flag the HIV positive status by using Z21 as a contributory cause. Some mortality systems do not accept Z codes, in that case R75 could be used instead.

O98, maternal infections and parasitic diseases classifiable elsewhere but complicating pregnancy, childbirth and the puerperium, might look like an alternative. However, both Z21 (which includes HIV positive) and R75 (laboratory evidence of HIV) as well as B20-B24 are excluded from O98. My suggestion is to use O15.9 for the underlying cause.

Following the same line of argument I would code the second case to septicaemia in newborn, P36.9.

In the third case I would use P01.1, fetus and newborn affected by premature rupture of membranes, for the underlying cause.

And then to the issue on how to record the HIV-positive status. Using Z21 or R75 might be overcoding the cases, but I think that would be a lesser evil than losing the information on the HIV status. So my suggestion is to use Z21 as a contributory cause for all three deaths. Some mortality systems do not accept Z codes, if so R75 could perhaps be used instead.

2004-02-10_01 Myocarditis due to viral infection

(UC Chapter I, IX)

[Question from Elaine Tower, England:]

We have a death here which we would like to know what your thoughts are on the Index routes.

1a Cardiac arrhythmia
1b Myocarditis
1c Viral Infection

Part II Verdict Natural Causes

The Sequence is acceptable, we would like to link B34.9 to I51.4 to give
Viral myocarditis  = I40.0.
The tables do not link.
The Index shows Myocarditis
               - virus, viral NEC I40.0

What are other countries doing? Thank you for your assistance

Canada /Patricia Wood/:
I agree with Elaine, Dr. Laurenti and Lars Age ... the underlying cause of death should be coded to ICD-10 code I40.0, infective myocarditis. 

Although the ACME decision tables do not link B349 and I514, I think that this particular circumstance is dealt with in the Intention of Certifier instructions in NCHS Instruction Manuals Part 2a Section III 3.a.(1)(b)(i) (pages 70-71) and Part 2b Section III 3.a.(1)(b)(i) (page 84).  Coders are instructed that, "when an infectious or inflammatory condition is reported and ... a condition classifiable to B34 is reported as the only entry on the next lower line ... if a single code is provided for the infectious or inflammatory condition modified by the condition classified to B34, use this code.  Do not assign a separate code for the condition classifiable to B34."

I checked what MMDS would assigned as the underlying cause of death in this case - I400 was selected!

Sao Paulo Classification Centre /Ruy Laurenti/:
In relation to the questions from Elaine, in the first case we code as viral myocarditis I40.0. We accept the "viral" as a qualification of the

Sweden /Lars Age Johansson/:
Like Ruy, we would code this to viral myocarditis. The linkage tables in Vol 2 do not have anything on B34.9 and myocarditis, but specificity (Rule D) certainly applies - "[Rule D] will often apply when the general term becomes an adjective, qualifying the more precise term" (Vol 2, p 43).

2004-02-10_02 Code for avian influenza

(UC Chapter I, X)

[Question from Moriyo Kimura, Japan:]

As you already know, bird flu became a lethal disease in Asia. We have been discussing ICD-10 code for it and decided to use J10 with a special footnote. Have you discussed or already allocated the code for the disease?

Canada /Patricia Wood/:
I agree with Dr. Kimura's suggestion to code avian or "bird" flu to J10.- Recently there was an outbreak of bird flu at a Canadian poultry farm but no human cases contracted/reported at this point. 


Germany /Robert Jakob/:
There have been no cases and no discussion either regarding the coding of "bird flu".


Sao Paulo Classification Centre /Ruy Laurenti/:
Regarding the "bird flu", we have not yet discussed the code for this disease.

Sweden /Lars Age Johansson/:
We have not had any cases so far, but I agree with Moriyo. J10 (Influenza due to identified influenza virus) seems very appropriate

2004-02-17_01 Senility an acceptable cause of death?

(Rule A)

[Question from Sao Paulo Classification Centre /Ruy Laurenti/:]

The Brazilian Center has been contacted by clinical physicians and epidemiologists that don't agree with the fact that the ICD norms/rules/guides don't accept senility (R54) as the underlying cause of death. They don't agree because there have been cases of elderlies (80-85 years old or older) that died and there aren't clinical reasons for it. We know there is "atherosclerose", but there aren't enough complications from it to justify the death. The same may happen with arterial hypertension and even diabetes.
The Brazilian Center is iniciating a study that consists in interviewing physicians that declared senility as the underlying cause of death and asking their oppinions. We would like to know the oppinion of the other members of the Mortality Forum and clinical physicians, specially geriatrics, on the subject.We haven't found enough references on the subject so we would appreciate if somebody could send us some.


Germany /Robert Jakob/:
I apologise, Ruy. I obviously have misunderstood your question.
From my point of view I really would accept senility as cause of death and I follow completely the argumentation the medical colleagues Lars is
referring to.

USA /Harry Rosenberg/:
For what it's worth, there is a literature on the subject. Go back many years ago, perhaps 20, for an interesting article in the Journal of the American Medical Association arguing strongly for "senescence" as a cause of death. Currently, in the U.S. Joanne Lynne, a respected gerontologist, in discussions with me about a project on palliative care in which we are involved, also feels that senescence -- a running down of the body -- is a legitimate description of the cause of death for many of her patients. The exact citation to the article on "senescence" as a cause of death for the elderly is as follows: Kohn RR. Cause of death in very old people. Journal of the American Medical Association 247(70): 2793-7. 1981.
I should add, as a caveat, that I believe that senescence and senility should be regarded as a default when other conditions have been ruled out, paralleling the guidelines used in the U.S. for Sudden Infant Death Syndrome. First, exclude everything else.


Sweden /Lars Age Johansson/:
For what it is worth, it is certainly my impression that many physicians regard senility as an acceptable cause of death. The patient might have a couple of diseases, but nothing serious enough to explain the death. Also, the death seems more to be due to a general decline than to complications of a particular disease. I have had a quite a few phone calls from doctors who want to know if they can put just "senility" on the death certificate, and I always say that they should write "senility" rather than just making something up.


Sao Paulo Classification Centre /Ruy Laurenti/:
Dear all, we received the answer from Robert Jakob about senility. Thank you Robert, but you answered what everybody knows! What we really want to know is the following:
a) can Senility in some cases clinically be accepted as cause of death? We are not talking about "death certificate" or "rule A".
b) we would like to receive references of papers regarding "senility" from the medical/clinical point of view.


Canada /Patricia Wood/:
I think that the clarification Robert makes is a very important one. Senility and old age (i.e. conditions coded to ICD-10 code R54) are acceptable underlying causes of death UNLESS reported with other conditions that are not ill-defined.  I think that this may be a certification issue rather than a classification issue.  If, in the opinion of the certifier, the atherosclerosis, hypertension or diabetes neither caused nor contributed to the death it shouldn't be certified.

Luxemburg /Monique Differding/:
I agree with Robert and must say that we have a lot of those cases. If there are no ill-defined  causes or other conditions on the certificate , we will use senility as U.C


Germany /Robert Jakob/:
Modification Rule A says:
"Where the selected cause is ill-defined and a condition classified elsewhere is reported on the certificate, reselect the cause of death as if the ill-defined condition had not been reported,...".

This implies if there is no other condition classified elsewhere reported on the certificate it would be correct to keep senility as UC.

2004-03-02_01 Code for sclerosing mesenteritis

(UC Chapter XI)

 [Question from Patricia Wood, Canada:]

Have any of you ever been asked where to code sclerosing mesenteritis? I had a brief (and overwhelming) look on the internet and noticed that it seems to be related to, or synonymous with, mesenteric panniculitis and retractile mesenteritis.  Neither of these terms are indexed either!  I am thinking about K65.8, Other peritonitis based on a very weak logic that if the mesentery is part of the peritoneum and it is inflamed ... I would be interested in knowing what you think.


Germany /Robert Jakob/:
We have already indexed "Mesenteritis" to K65.9. We had no question for the "Sclerosing Mesenteritis" yet. From the literature I see that there is an idiopathic one. Other forms are associated with specific diseases. So the term "sclerosing" to me looks not very "specifiing " to me and I would keep K65.9 for the "Sclerosing Mesenteritis". In fact peritoneum and mesenteriom are not the same, but as for K66, in K65 is mentionet mesenteriom, too (e.g. K65.8 Mesenteric fat necrosis). I suppose K68.8 is a typo unless you don't have it in a Canadian adaptation of the ICD-10.

2004-03-02_02 Complication of laparoscopic surgery

(UC Chapter XX)

[Question from Lilja Sigrún Jónsdóttir, Iceland:]
Dear members of the Mortality forum. We have a certificate here to admit to the forum:
A female at 41 yrs
Part I
1a Fat embolism syndrome
1b Small bowel perforation with complications
1c Lysis of adhesions during legal sterilization
First operation was June 14th, where the small bowel perforation was mended, on the 23. reoperated due to abdominal abscesses.

When coding misadventures during surgery, it seems ICD10 does not specify complications of laparoscopic surgery? It would seem important to be able to identify such instances separate from other surgery since this technique is taking over an ever increasing part of surgical operations. I have suggested the following coding myself: T791/T812/Y600/Z302, UC Y600
Please comment.
[Suggested coding:]
  Underlying  Multiple causes
Canada     N736     T791/T812/Y600 N736
   Y600     T791/T812/Y600
Iceland  Y600  T791/T812/Y600/Z302
Sweden N736   T791/T812/Y600/N736


Canada /Patricia Wood/:
In Canada, the "Z" codes are not valid for use in mortality classification, neither as underlying cause codes nor as multiple cause codes. ICD-9 Rule 12 did not survive the revision process, so in ICD-10 misadventures to patients during surgical or medical care are treated in the same way as abnormal reactions or later complications - the condition necessitating the treatment, if stated or implied, is classified as the underlying cause of death.
Is it reasonable to infer, from this certificate, that the small bowel perforation occurred during the lysis of adhesions (during the sterilization) making the adhesions the condition necessitating the surgery? Otherwise we are left with the sterilization for which there is no real "condition" requiring treatment, which would lead to Y600 as the underlying cause of death.
Is the description of the sterilization as "legal" significant?

Sweden /Lars Age Johansson/:
I agree with Patricia, and would go for the first alternative. As the certificate is written, it seems that the adhesions were present before the sterilization, and that the surgical treatment of the adhesions caused the complication rather than the sterilization itself. Lilja is certainly right that it is difficult to identify laparoscopic surgery from ICD codes, and that monitoring complications of surgery is important. The ICD contains codes for some surgical techniques, but given the enormous range of techniques I wonder if it would better to use ICD codes to identify the complication, and a code from a procedure classification to identify the surgical technique? Death certificates seldom contain enough information on the kind of operation performed, but procedure data could perhaps be obtained from hospital discharge registers.

2004-03-23_01 Newborn, cause of death undetermined

(Rule P3, UC Chapter XVI)

[Question from New Zealand /Claire Nicoll/:]

Please can we have some feedback on the following case:
Newborn infant - 7days old breastfed by mother in single bed -both fall asleep. 2 hours later baby is discovered dead.  The coroner rules that the cause of death is undetermined.
Using rule P3 we are directed to use P95 as cause of death -however the description of this code, "fetal death of unspecified cause" implies stillbirth which is clearly not the case here. Using the index,  "death, neonatal" we are directed to the preferable P96.8. Is it appropriate to apply rule P3 or should we go with instinct and use P96.8?

Sao Paulo Classification Centre /Ruy Laurenti/:
In Brazil we do not use a different death certificate for perinatal deaths. In this case there is a strong possibility that the death occurred by asphyxia due to due accidental suffocation in bed (mother's body). However the coroner informs that the cause of death is undetermined. We have doubts in relation to the coding: R95 or P96.9?


PAHO /Margaret Hazlewood/:
R95 may be considered quite appropriate for this death. Why? P96 refers to "Other CONDITIONS originating in the perinatal period."  Is death considered as a "condition" originating in the perinatal period???


Canada /Patricia Wood/:
In Canada, we do not have a separate perinatal death certificate so we do not use the perinatal death coding rules, but surely ICD-10 code P95, Fetal death of unspecified cause, is not appropriate as an underlying cause of death code for any "liveborn" infant.

In this case I would code the cause of death as P969, Condition originating in the perinatal period, unspecified.  The "perinatal period" is defined (Volume 2, page 130) as ending seven completed days after birth and this baby is said to be 7 days old.  In this same scenario with an older baby, I think that the only possibility would be R95, Sudden Infant Death Syndrome which includes sudden death of nonspecific cause in infancy (or at least it did in ICD-9!)  


Nordic Classification Centre /Martti Virtanen/:
There are two important points that this question rises:
1. Death during first weeks of life is rare in Sudden Infant Death Syndrome, typical age is 4-16 weeks. Therefore, the postmortem examination should be especially careful in this kind of cases. The child could for example have had an inborn error of metabolism which would be very important to know before a possible next pregnancy.
2. Under rule P3 the example 5 is difficult and the example 6 impossible to understand. Fetal period ends at birth and therefore a fetal death cannot occur after birth - at least fetal death at 2 days of age must be an error. In addition to code P96.9 the rule should also discuss the possibility of using R95 - it is definitely a possibility in this kind of situation. As for the coding I think R95 is the only possibility, if no more
information is available.

Sweden /Lars Age Johansson/:
As far as I understand, the code P95 in example 6 (Vol 2, p 95) is erroneous. The explanatory text says that the death is coded to "unspecified perinatal cause", but the code for unspecified perinatal cause is P96.9 and not P95. P96.9 is also the code given by the text of Rule P3.
Since it is not clear from Claire's description of the case whether and autopsy had been performed and other causes of death ruled out, I would hesitate to use R95.


Jordan /Majed As'ad/:
Dear members of the Mortality Forum,
I think this case must be coded as R95 sudden infant death syndrome (sudden and unexpected death ofan infant who had previously been apparently well, and which is unexplained by careful posmortem examination).


2004-03-23_02 Subdural hemorrhage due to Parkinson's disease

(Rule GP)

[Question from Norway /Gunvor F. Østevold/:]

How are you dealing with a certificate like this:
65 year old male
1a) subdural haemorrhage, operated
 b) -
 c) Parkinson's disease

According to Volume 2, 4.2.2 (m) (s68) epilepsy is the only disease which can cause any accident, but patients with Parkinson's disease do fall a lot so I think subdural haemorrhage could be caused by Parkinson's disease?

Sao Paulo Classification Centre /Ruy Laurenti/:
For the case from Norway/Gunvor the code attributed to the underlying cause of death in the case is I62.0, according to volume 2, item (m) 4.2.2.


Australia /Margaret Campbell/:
A quick literature searched showed that:

Subdural haemorrhage, usually occurs as a result of minimal trauma, or as a result of cerebral atrophy (in the elderly), which causes a gradual widening of the subdural space leading to rupture of the bridging veins http://www.fleshandbones.com/readingroom/pdf/463.pdf.

It would appear that there is a higher risk of cerebral haemorrhage in patients with cerebral atrophy as with Parkinsonism.  This would support the decision to code to G20 as UCOD with I620 however another scenario that occurred to me on viewing this certificate was that the subdural haemorrhage could be a complication of surgery to control Parkinson's eg, Pallidotomy or Thalamotomy.  The age of the patient might indicate whether this scenario is more or less likely.  For this reason I would seek clarification from the clinician prior to coding.


Canada /Patricia Wood/:
Is this really a case of a disease causing an injury ... subdural hemorrhage is indexed to ICD-10 code I62.0 and, in the absence of an indication of trauma, that's how I would code it in this case.  Then applying the General Principle, I would select G20, Parkinson's disease as the underlying cause of death.


New Zealand /Claire Nicoll/:
In the first instance we would ascertain from the certifying doctor if the cause of the subdural haemorrhage was traumatic (we would not assume this if surgery was mentioned). If no further information was available we would assume that the haemorrhage was spontaneous and code I620. The only medical condition we accept as the cause of an accident is epilepsy.

Sweden /Lars Age Johansson/:
When the "Mortality Reference Group" started back in 1998, one of the first issues we discussed was whether sequences like this one should be accepted. There are several cases where you could easily imagine that an injury is caused by a disease, such as if someone standing on a ladder gets a heart attack and falls from the ladder, or someone driving a car suddenly gets a stroke and because of that loses control over the car. However, we finally decided to leave things as more or less as they are, so with a few exceptions a disease should not be accepted as the cause of an accident. The exceptions are, according to the updated version of Vol 2, an accident reported as due to epilepsy, a fall reported as due to a disorder of bone density, asphyxia due to aspiration of mucus, blood or vomitus reported as due to diseases, and aspiration of food reported as due to a disease which affects the ability to swallow. The main reason why the MRG did not want to allow for more sequences of this is kind was that, in most cases, information on the injury is more important from the preventive point of view than information on the underlying disease.

2004-03-23_03 Code for long QT syndrome

(UC Chapter IX)

I would like to have your opinion about a question received by Forum-CIE from Panama:

How to code
Long QT Syndrome


Canada /Patricia Wood/:
I checked NCHS ICD-10 Volume 3 and long QT syndrome is indexed to R94.3 Abnormal results of cardiovascular function studies.  Romano-Ward syndrome is not indexed, but Jervell and Lange-Nielson syndrome is indexed to Q99.8 Other specified chromosome abnormalities.  I also checked each of these causes of death in MMDS Version 2004.01 but I don't think that any of them are in the MICAR dictionary as they all rejected with code U00.   

PAHO /Roberto Becker/:
Thanks for the answers. In fact I circulated this question in three Forum's:
- Forum-CIE
- URC-Forum
- Mortality Forum

Please, see below a sumary table with all the opinions I got:                       

  SQTL S Jervell and  Lange-Nielsen S Romano-Ward
España Q24.8 Q24.8   Q24.8
Panamá I45.8  I45.8    I45.8
México I45.9 Q99.8  Q24.6
Cuba    I45 or I49    
New Zealand I49.8   Q99.8  Q99.8
Deutschland  I45.8  Q99.8 Q99.8
UK    I45.8  Q... I45.8
Canada   I47.2     Q99.8   Q99.8
Brasil   I45.8 Q99.8 Q99.8
Australia I45.8/I47.2 I45.8     I45.8
Mortality Forum in 2000:
Australia   I49.8
Canada R94.3
Deutschland I45.8
Slovenija I49.0 or I49.8
Sverige I45.8
USA R94.3

As a provisional solution I'm recommending in Forum-CIE to code as Michael Schopen and Ruy Laurenti mentioned (See above Deutschland and Brasil).
I'm proposing to the URC exactly what was mentioned by Julie Rust in the URC-Forum: "this question should be put forward as a proposal for 2004 (linked to URC:0208) in order to kick start the work in mapping the nomenclature lists to ICD-10."


New Zealand /Claire Nicoll/:
We have always assigned I498, "Other specified cardiac arrhythmias" for Long QT Syndrome where no other information has been available.  For the variants Jervell and Lange Nielsen Syndrome and Romano-Ward Syndrome - both are genetic disorders without specific syndrome codes so we would probably assign Q998, "Other specified chromosomal abnormalities".

Sweden /Lars Age Johansson/:
We have had the question on "long Q-T syndrome" once before in the forum (2002-09-30 Q2), but we did not agree on which code to use. The suggestions were:
Australia I498
Canada  R943
Germany I458
Slovenia  I490 or I498
Sweden   I458
USA    R943

2004-03-23_04 Code for Jervell Lange Nielsen Syndrome and Romano-Ward syndrome

(UC Chapter IX)

I would like to have your opinion about a question received by Forum-CIE from Panama:

How to code
The two variations:
a) Jervell y Lange Nielsen Syndrome
b) Romano-Ward Syndrome


New Zealand /Claire Nicoll/:
We have always assigned I498, "Other specified cardiac arrhythmias" for Long QT Syndrome where no other information has been available.  For the variants Jervell and Lange Nielsen Syndrome and Romano-Ward Syndrome - both are genetic disorders without specific syndrome codes so we would probably assign Q998, "Other specified chromosomal abnormalities".

2004-03-30_01 Code for multiple system atrophy

(UC Chapter VI)

[Question from Belgium, Josiane Mornie:]

How do you code " Multiple system atrophy "? G23.8? G31.9? G90.3? Thanks for answer !

Australia /Margaret Campbell/:
Originally called the "Shy-Drager Syndrome" this complex syndrome is currently referred to as "Multiple System Atrophy" or MSA http://www.ndrf.org/MSA.htm I would code to G90.3 to be consistent with original meaning.


Sao Paulo Classification Centre /Ruy Laurenti/:
Multiple System Atrophy: Why to use a G code? How we know that is a nervous system disease? For us is "multiple system" and not "nervous system". For us the code is R68.8.

Sweden /Lars Age Johansson/:
We use a G code since the vast majority of citations we have found describe MSA as a neurological disorder (encompassing Shy-Drager syndrome, striatonigral degeneration, and olivopontocerebellar atrophy). However, I agree with Ruy that sometimes it seems more likely that the certifier has a general degeneration of body systems in mind, rather than a specific neurological condition.


England /Elaine Tower/:
Here in England we code "Multiple System atrophy" to R68.8 as multi-organ failure where it is mentioned on it's own. Only where it states Shy Drager on the certificate after the words "Multiple System atrophy" would we use the code G90.3.

Saxony /Birgitt Göldner/:
We think that the correct code for "Multiple system atrophy" is G90.3 as in the "Systematical Index".

Sweden /Lars Age Johansson/:
Like Saxony, we code "multiple system atrophy" to G90.3. However, we use R68.8 for "multiple organ failure" and similar, less precise terms.

2004-04-20_01 External cause code for "smoke inhalation"

(UC Chapter XX)

[Question from Canada /Patricia Wood/:]

Every year we have certificates with "smoke inhalation" as the cause of death, but no other detail provided.  When possible we seek specification of the source of the smoke, however we are often faced with having to code unspecified "smoke inhalation" as the underlying (external) cause of death.  How best to code this?  Inhalation, smoke is indexed as a nature of injury to T598, Toxic effect of other specified gases, fumes and vapours.  It is not indexed as an external cause specifically, but Inhalation, toxic gas directs one to the Table of drugs and chemicals where smoke NEC is indexed to X47, Accidental poisoning by and exposure to other gases and vapours.  Alternately, dare I suggest that, "where there's smoke, there's fire" and look at the block, X00-X09, the title of which is Exposure to SMOKE, fire and flames, and specifically X09, exposure to unspecified smoke, fire and flames as a possibility for classifying "smoke inhalation."  What do you think?



Luxemburg /Monique Differding/:
In Luxemburg , for accidentaly smoke inhalation we code like Elaine ( X09.. ), and in case of unknown we do the same as Ruy ( Y 26.. ). In this case it is very important to know if the intention was an accident or suicide or aggression or even unknown.


Cuba /Ana C. Mesa/: 
I think X479-T598.

Sao Paulo Classification Centre /Ruy Laurenti/:
In this case we does not know it is "accident or suicide". The Brazilian Center code as Y26.9.


England /Elaine Tower/:
Here in England if it is stated as only "smoke inhalation" with no other causes present we would use the following codes: T59.8 and X09.9 "Exposure to Smoke".

2004-05-12_01 Code for chest infection

(UC Chapter X)

[Question from Australia (NCCH) /Margaret Campbell/:]

As no code trail exists in the ICD-10 index for the term 'chest infection' and not all countries are using MMDS software I would be interested to know if the term 'chest infection' is universally classified to J98.8 Other specified respiratory disorders.

In Australia mortality coders classify 'chest infection' to respiratory tract infection NEC J98.8 as does the MMDS software, however morbidity coders classify 'chest infection' to J22 Unspecified acute lower respiratory infection.

1. Chest infections: Acute infections of pulmonary parenchyma associated with at least some symptoms of acute infection [ www.aic.cuhk.edu.hk/web8/chest_infections.htm ]

2. Chest infection: another name for Mediastinitis which involves inflammation of the mediastinum, the cavity between the lungs. Mediastinitis is classified to J98.5
[ www.nlm.nih.gov/medlineplus/ency/article/000081.htm ]



Malta /Kathleen England/:
In Malta we code Chest infection as J22. Most doctors here think of chest infection in the same way as they think of pneumonia or lower respiratory tract infection. If we coded chest infection to J98.8 we would lose a number of cases of death due to respiratory infections since J98.8 is quite non specific.


Sweden /Lars Age Johansson/:
Swedish does not have any equivalent to "chest infection", so we have not been forced do make a decision on how to code the term. However, an internet search showed that the term is used of a wide variety of conditions like acute bronchitis, pneumonia, mediastinitis, unspecified lower respiratory tract infections, TB, and SARS. In view of that, J98.8 would seem the "safest" code, when no additional information can be obtained from the certifier.

2004-05-12_02 Code for Baller-Gerold syndrome

[Question from Cuba, forwarded by Roberto Becker, PAHO:]

How do you code Baller-Gerold Syndrome ?

2004-05-12_03 Cholecystitis with perforation, nephopathy in Part II

(Rule GP,3)

[Question from Germany /Michael Schopen/:]

Ia) Septic multiple organ failure
b) Peritonitis
c) Cholecystitis with perforation
II Nephropathy, arterial hypertension



Sweden /Lars Age Johansson/:
I would read line 1c as "cholecystitis causing rupture of gallbladder" and then apply the General Principle, since cholecystitis may very well explain both rupture of gallbladder, peritonitis and septic organ failure. As far as I can see cholecystitis is not an obvious consequence of either nephropathy or arterial hypertension, so I would select unspecified cholecystitis (K81.9) as the underlying cause of death.

2004-05-12_04 "History of" reported as originating cause

(Rule GP,A)

[Question from Germany /Michael Schopen/:]

Ia) Senile cachexia
b) History of small pelvic surgery
c) History of mastectomy


Sao Paulo Classification Centre /Ruy Laurenti/:
The Brazilian Center agrees with Lars. For us the u.c. is R54.


Sweden /Lars Age Johansson/:
In Sweden, we take "history of" and similar expressions to mean that the condition was cured and did not cause the death. If there is a suitable "history of" code (Chapter XXI) available we will use it as a multiple cause of death, but in this case we do not know why the operations were performed. It is quite likely, of course, that the operations were carried out because of malignancies, but there is no indication that they were active at the time of death and therefore I wouldn't code them. The only remaining candidate for underlying cause of death is senile cachexia, R54.

2004-05-12_05 Neoplastic disease and mastectomy

(Rule D)

[Question from Germany /Michael Schopen/:]

Ia) Apoplexy
b) Neoplastic disease
II Mastectomy



Sao Paulo Classification Centre /Ruy Laurenti/:
The Brazilian Center agrees in part with the reasoning from Lars. However for us the information of "mastectomy" is a indication of breast cancer and, in this case, "neoplastic disease" is a synonym of "metastasis". By application of Rule 3 breast cancer is the u.c.

Sweden /Lars Age Johansson/:
Since the General Principle applies, I would select "neoplastic disease" as the underlying cause of death. Assuming that "neoplastic disease" means "malignant neoplastic disease" in Germany (as it does in Sweden) I would use C80 for the underlying cause of death. It could be argued, of course, that the mastectomy mentioned in Part II indicates that the malignancy was a breast cancer, but since it is quite possible that the patient had breast cancer previously but now died from a different malignancy, I would stay with C80.

2004-05-12_06 Multiple neoplasm due to HIV

(Rule GP)

[Question from Germany /Michael Schopen/:]

Ia) Multiple neoplasms
b) HIV



Cyprus /Pavlos Pavlou/:
Does Lars Age's interpretation of the above statement mean that HIV may be assumed to be a cause of any malignant neoplasm?
If this is the case would it not conflict with Vol 2, p 39, Assumed direct consequences of another condition:
" Kaposi's sarcoma, Burkitt's tumour and ... should be considered to be a direct consequence of HIV disease, where this is reported. No such assumption should be made for other types of malignant neoplasm."?

Also, does this last statement not conflict with the existence of code B21.8, HIV disease resulting in other malignant neoplasms ?

To simplify my question with two examples, what would the underlying cause of death be for a certificate like

I a) Carcinoma of lung
  b) HIV

and what would it be for a certificate like

I a) Carcinoma of lung

Sweden /Lars Age Johansson/:
Thanks for giving me the opportunity to clarify my reply! When commenting on Michael's example, I was only speaking of sequences reported in Part I of the death certificate. When a malignant neoplasm is reported in Part I as caused by HIV, then the sequence should be accepted. Consequently, in the case Michael submitted to the Forum B21.7, HIV disease resulting in multiple malignant neoplasms, is the underlying cause. Similarly, B21.8, HIV disease resulting in other malignant neoplasms, is the underlying cause in the first example provided by Pavlos. In both these cases HIV has been mentioned in Part I of the death certificate, and as the cause of the malignant neoplasm. However, when a malignant neoplasm is reported in Part I and HIV is mentioned in Part II, HIV should be considered the underlying cause only if
the neoplasm is Kaposi's sarcoma, Burkitt's tumour, or another neoplasms coded to C46 or C81-C96. As always, the requirements for selecting an underlying cause from Part II are far more rigorous than for accepting a sequence reported in Part I. Accordingly, the underlying cause in Pavlos's second example is C34.9, malignant neoplasm of bronchus and lung.


Sweden /Lars Age Johansson/:
According to Vol 2, p 68, a malignant neoplasm may not be due to any other disease - except HIV. This means that the General Principle applies here, and HIV is the originating cause. I would code the case to B21.7, HIV disease resulting in multiple malignant neoplasms (provided that "neoplasm" on a German death certificate means "malignant neoplasm", as it would in Sweden).

2004-05-25_01 Code for viral gastritis

(Chapter I, XI)

Question from England /Elaine Tower/:

Here in England we are having a problem coding a death and we would welcome any comments.

A 91 year old female with
1a Viral Gastritis
Part II Alzheimer's disease, Rheumatoid arthritis

Viral Gastritis is getting allocated the code B33.8. Do you think this is the correct code? We were thinking that it may be A08.4 - viral intestinal infection, unspecified.



Cuba /Ana C. Mesa/:
I think K296 and additional code B978. A084 is "viral Intestinal infection", here is "gastritis".

Sweden /Lars Age Johansson/:
The Mortality Reference Group met two weeks ago and discussed, among other things, Elaine's question on infectious gastritis. We finally arrived at a solution very similar to the one suggested by Ana: use an appropriate code in K29, and also use an additional code from B95-B97 to identify the infectious organism.

2004-06-15_01 Code for Escobar syndrome

(Chapter VI)

[Question from Belgium /Josie Mornie/:]

We have a girl - 2,5 years - with  a syndrome of Escobar. Do you know this? And how can we code it? (= a kind of encephalopathy ...)
[Suggested coding:]
Australia   Q99.8
Cuba    Q87.8
Jordan      Q18.3
Sao Paulo ClC  G93.4


USA /Julia Raynor/:
I agree with Australia and Cuba and would code Escobar syndrome to Q878. As Margaret and Ana stated, synonyms are pterygium and multiple pterygium syndrome. We code those terms to Q878 also.

Australia /Maragret Campbell/:
I now agree with Cuba /Ana C. Mesa/ and would code to Q87.8.


Australia /Margaret Campbell/:
Synonyms: Familial pterygium syndrome, pterygium syndrome.Multiple pterygium syndrome.

A rare disturbance manifested by orthopaedic and cranial anomalies. Characteristic features include short stature, craniofacial anomalies, joint contractures, vertebral fusion anomalies, rocker-bottom feet, and pterygia of the neck, antecubital, digital, poplietal, and intercrural areas. Abnormalities of the head usually consist of epicanthal microcephaly, skin folds, long philtrum, antimongoloid palpebral slant, low-set ears, pointed and receding chin, ptosis, down-turned angles of the mouth, cleft lip and palate, and hemangiomas of the forehead. Associated anomalies may include rib defects, scoliosis or lordosis, vertical talus, cryptorchism, hypoplastic labia majora, and mental retardation. Most cases are transmitted as an autosomal recessive trait, some as a dominant trait, and a few are sporadic.

I would code to Q99.8 Other specified chromosomal abnormalities.

This appears to be transmitted as an autosomal trait, therefore I looked up syndrome, autosomal in the index (Q99.9).  As the syndrome is specified on the certificate but not classified.

Cuba /Ana C. Mesa/:
This Syndrome is similar to (family) Syndrome Pteregium multiple lethal, ICD-9 (759.8), ICD-10 (Q87.8). (Hygroma bladder, hypoplasia of lung, hypoplasia of heart, contracture of joint, malformations of face and other malformations) - Syndromes affecting multiple systems.

Jordan /Majed Asad/:
In the case of syndrome of Escobar I would code the cause of death as Q18.3.
Sao Paulo Classification Centre /Ruy Laurenti/:
In the Brazilian Center we never saw the diagnosis "Syndrome of Escobar". As there is the information that is a kind of encephalopathy, for us the code is G93.4.

2004-06-15_02 Code for haemophagocytic syndrome

(Chapter III)

[Question from Luxemburg /Monique Differding/:]

Please, could you help me in this difficult case!

Child of 8 months
a) Bronchiolitis  respiratory infection after bone marrow transplantation
b) Asphyxia
c) Hemophagocytic syndrome

I think that the code D762 will not be the right one?


Cuba /Ana C. Mesa/:

Vol 3:
- hemophagocytic, infection-associated D762

- hemophagocytic (infection-associated) D762
Is different condition.

Other possibility: R71. I think D76.2 is possible.

Jordan /Majed Asad/:
In the case of child of 8 months D76.2 will be the right one.

Sao Paulo Classification Centre /Ruy Laurenti/:
The u.c. is D76.2.

2004-06-15_03 Use of codes in P00-P04

(Chapter XVI)

[Question from Malta /Kathleen England/:]

In Malta we do not have a Perinatal death certificate and rely on the General form of death certificate for perinatal deaths. Because  of this I sometimes am getting confused as what should I code as cause of death. E.g., perinatal death (stillbirth) due to placental abruption is coded to P02.1. However, the ICD rules and guidelines categories P00-P04 to maternal conditions affecting fetus and therefore I am not sure if I can use it as a underlying cause of death of the fetus.



Australia /Margaret Campbell/:
The gestational age of the fetus is very important in coding these cases.  If the terminology used is stillbirth then the assumption is that the fetus was viable at the time of death and your code selection is correct.  The code range P00-P96 relate to conditions originating in the perinatal period even though death or morbidity occurs later. The perinatal period starts as the beginning of fetal viability (28 weeks gestation or 1000g) and ends at the end of the 7th day after delivery. Perinatal deaths are the sum of stillbirths plus early neonatal deaths.

Cuba /Ana C. Mesa/:
Not to use P00-P04 for the "Main disease or condition in fetus or infant" is correct, but it can be used for underlying cause of death, of  any age.

Sao Paulo Classification Cente /Ruy Laurenti/:
For us the u.c. of death in the case is P02.1.

2004-06-22_01 Hyponatraemia due to drinking 7-8 litres of water

(Chapter IV, XX)

[Question from New Zealand /Claire Nicoll/:]

Please could we have some suggestions for the following case:

A 41 year old man attended a life mastery course which included body cleansing - drinking 7-8 litres of water - he lost conciousness and died due to hyponatraemia. As the hyponatraemia was self-induced (but accidental), is E871 the appropriate code?

Slovenia /Jozica Selb/:
Dear colleges, I would add to E887 also X58 because it was an accident and not an consequence of medical treatment or any other.


Sweden /Olafr Steinum/:
Being not a mortality coder but  a morbidity  physician with knowledge of clinical coding, I am not familiar with  the more sophisticated rules in morbidity coding. But: I totally agree with prof Laurenti. This is a case of death due to an external cause.  E87.-  is only an additional description.

Sweden /Lars Age Johansson/:
There are no special rules we could invoke in this case, as far as I know. After reading the correspondence, I agree that E87 is rather a description of the effect than of the cause. However, I'm not sure about the best code for the excessive drinking. The closest thing to over-drinking I have been able to find is polydipsia, which is indexed to R63.1. But if we use R63.1 for the drinking and E871 to describe the effect, then E871 would be the tabulated underlying cause since R631 is considered ill-defined. The consequence would be that we lose all information on the actual cause of
this death. That is a great pity, considering that this is an exceptionally preventable cause. Perhaps we need a specific code for deleterious dietary habits?

Canada /Patricia Wood/:
I am inclined to agree with Dr. Laurenti that the underlying cause should be classified to an external cause code.  Neither the fluid overload nor the hyponatremia resulted from a disease process but rather from an "external" cause, the excessive water intake.  Statistically, maybe this death should be included with other accidental deaths rather than with deaths from "natural" causes.
England /Elaine Towers/:
In a routine death, we would code Hyponatremia to E87.1 with Fluid overload to E87.7 and give an underlying code as E87.7. But if the scenario was a coroner's verdict the codes would be E87.1 with Y546 being the code for adverse effect in correct usage (if he was following strict guidelines) OR E87.1 with T503 and X448 and the Underlying codes would be T503 and X448 pointing to poisoning / Intoxication of water. (If he had gone against advice).


Sao Paulo Classification Centre /Ruy Laurenti/:
Very interesting the case you sent. We never see a case like this:
"intoxication"(!!) by water with a consequent hyponatraemia. For us is a external cause accidental. As the ICD-10 does not have a specific code for this we think that the appropriated code is X58._ u.c. and E87.1 as associated code.
Ruy laurenti
Jordan /Majed Asad/:
What about E87.7  Fluid overload?
Sweden /Lars Age Johansson/:
We have had a couple of similar cases, usually described as "water intoxication". Like our Jordan colleague, we have classified them to E87.7.