Letters

In reply: Whether Hospital in the Home (HIH) care is suitable for managing patients with community-acquired pneumonia (CAP) depends on what is considered an appropriate use of resources. Overall, we see relatively few indications for treating CAP patients with parenteral antibiotics via HIH, as, in our experience, most patients who do not need supplemental oxygen and are well enough to be treated at home are usually also well enough to be treated with oral antibiotics. If they are not well enough to take oral antibiotics, then admission to hospital as an inpatient is generally appropriate. The occasional exceptions to this are selected patients in nursing homes (where around-the-clock supervision is available if required) and some patients with CAP caused by pathogens like Pseudomonas or Acinetobacter who benefit from longer treatment courses and may not have the option of oral antibiotics. Furthermore, a report submitted to the Victorian Department of Human Services regarding HIH care of CAP patients at a number of Melbourne HIH units identified significantly worse outcomes at some centres, mainly related to inappropriate patient selection. Notable, but fortunately rare, cases included some patients with pulmonary embolism incorrectly diagnosed as CAP. Given that between 20% and 50% of patients given a diagnosis of CAP in the emergency department do not have pneumonia confirmed by a radiologist,1-3 the ability of busy emergency department doctors to select patients appropriately is definitely a concern. Thus, HIH treatment of CAP patients may be appropriate occasionally, but very careful patient selection is vital. In response to Bradley, the statement that there have not been any failures in treating pneumococcal CAP with penicillins refers to microbiological failures due to antibiotic resistance. Although several studies have suggested that combination therapy may reduce mortality from bacteraemic pneumoccocal infections, all of these have been retrospective observational studies. Thus, they lack the ability to control accurately for potential confounding variables such as disease severity, patient or family wishes, pre-morbid quality of life, or “not for resuscitation” status. Data are also lacking on co-infection with “atypical” pathogens such as Legionella. The immunomodulatory effects of macrolides, quinolones and tetracyclines on treatment response are also still being elucidated.4 We agree with the CAP treatment recommendations in the Australian antibiotic guidelines,5 which recommend dual therapy with a β-lactam antiobiotic plus either a macrolide or doxycycline for all patients who are not allergic to these drugs.

Patrick G P Charles · Paul D R Johnson · M Lindsay Grayson

To the Editor: The article by Jordan and Osborne1 highlights some of the key issues to be addressed if chronic disease self-management programs are going to be effectively incorporated into the Australian health care system, particularly in primary care. Although the National Chronic Disease Strategy2 recommends that self-management programs be integrated and supported at all entry points into the health care system, many of the self-management strategies and programs have been developed with little engagement of general practitioners and have not been integral components of primary health care. Jordan and Osborne highlight that, without the support of GPs, programs such as the Expert Patients Programme3 may have limited success. We recently completed a systematic review for the Australian Primary Health Care Research Institute to explore the evidence for managing chronic disease in primary care, with specific reference to the Australian health care system.4 The self-management programs found to be most effective were those that developed self-efficacy in relation to specific behaviours, such as diet and exercise for diabetes, rather than those that were more general. The combination of self-management support with delivery system design changes (such as multidisciplinary team care and follow-up) was effective in improving patient health outcomes for a number of chronic diseases. This highlights an important and developing role for practice nurses in chronic disease management. Given the burden of chronic disease in Indigenous populations, it is important to conduct more research on the role of self-management education and support in Indigenous communities, as there were few relevant studies identified in our systematic review. The funding available through the Australian Better Health Initiative5 will enable primary health care professionals, such as GPs and practice nurses, to receive self-management education training. Furthermore, to ensure that self-management support is embedded in primary care, we suggest that self-management support be included in care plans or annual cycles of care for conditions such as diabetes. Self-management support could also be incorporated into allied health services that are provided as part of a care plan.

Sarah M Dennis · Nicholas A Zwar · Iqbal Hasan · Mark F Harris

To the Editor: No one denies the right of insurers, or law firms acting for an interested party, to obtain information relating to a claim with due authority from the patient. After all, it’s fair enough for insurers to check the details of a claim. This system has worked well for decades and allows the insurance industry to operate reasonably and fairly. But a recent development gives cause for great concern. There is a trend among both insurers and lawyers to demand copies of a patient’s entire medical file. Not just details relevant to the claim, but the entire medical file, often extending for years before the relevant event. A cynic has suggested this is financially motivated, as a reasonable fee for copying and forwarding a file is likely to be less than the fee for reviewing the file and preparing a report. Another equally uncharitable explanation is that insurers are embarking on “fishing trips”, hoping to find grounds to mitigate a claim. I’m sure the industry could produce examples in which fishing trips have identified dishonest claims that might not otherwise have been detected. But I cannot justify sacrificing the privacy of many patients to expose the occasional fraudulent claim. In one case, one of my patients suffered a work-related injury that prevented her working as a private contractor for some months. The case was clear-cut, simple and straightforward. Yet the insurer refused to process her claim without receiving a copy of her entire file — including very personal details of a sexual assault nearly 20 years earlier, together with confirmation that she had contracted a sexually transmitted infection and details of her subsequent breakdown. For a vulnerable and very private person, it was a terrible ordeal to have this brought up and to have to sanction its disclosure to a claims officer. My financially strapped patient was held to ransom when the insurer advised that the claim would not be processed until the disclosure was authorised. How can the individual claimant ever stand up to a multinational insurer? And, in this case, appeals to the industry regulatory body were peremptorily dismissed. I support any stand by our profession against this unjustifiable invasion of privacy.

Bernard S Pearn-Rowe

Comment: Agents for a workers compensation authority commonly assert they have the right to see the patient’s entire medical history to assess whether the injury arose from work or from another past or current illness or injury. Two points need to be underscored. Firstly, patients have a right to waive their right to privacy. A consent for total disclosure, given when initiating a claim for compensation, could be invalid, as it is a consent given under (economic) duress — that is, payments will not commence unless the patient consents to full disclosure. A doctor may feel a duty to point out to the patient that full disclosure will reveal distressing matters from the patient’s past that the doctor believes are not relevant to the claim. The doctor may recommend that the patient seek legal advice on how to object to full disclosure. Whether the patient objects or agrees to total disclosure is a decision for the patient, not the doctor, with the help of legal advice. Doctors should not give that advice. Secondly, the legislation governing statutory compensation schemes in Australia gives extraordinarily broad powers to the relevant authority to demand information. For example, section 239 of Victoria’s Accident Compensation Act 1985 states that the Victorian WorkCover Authority (and its agents) may “require any person – to furnish the Authority with such information as the Authority requires . . . and may require the person to produce all books in the custody or under the control of the person relating thereto.” Doctors and medical records are not excluded from the broad sweep of section 239. Notwithstanding such clauses, a claimant can object directly to WorkCover about a request or demand to supply his or her entire medical history if the claimant believes there are matters not relevant to the claim that he or she does not wish to be disclosed. If that process fails, the claimant then has a number of legal avenues that can be pursued. Pearn-Rowe may be concerned at the David–Goliath imbalance of power between patient and insurer, but that does not justify omission of information because the doctor thinks it is not relevant. The Medical Defence Association of Victoria recently settled a case in which a member, asked to provide a “Personal medical attendant’s report” for a patient applying for a new disability insurance policy, omitted information about the patient’s history that would have affected the insurer’s assessment of the application. The patient had signed a consent for full disclosure. The insurer issued the policy, and a short time later the patient was diagnosed as suffering a major illness, the premonitory symptoms of which were described in the omitted information. Whether or not it was a deliberate omission was not the point — it was a negligent omission. In summary, the patient has a right to object to disclosure. If the patient chooses not to, the doctor has a legal obligation to comply with the literality of the patient’s signed consent for disclosure.

Paul Nisselle

To the Editor: “Out of the blue”, I received a personal message from a retired nursing sister of a children’s hospital, who had earlier written to the MJA1 describing “a cluster of neonates [who had] simultaneously turned blue” in the 1950s. Her letter to me followed one of mine in the Journal, describing methaemoglobinaemia (MetHgb) in infantile keto-acidosis, where the cyanosis responded rapidly to diabetic control alone.2 Rare as MetHgb is, it has previously been recognised as a presentation of infantile acidosis.3 The cause of the colour change in the neonate cluster “was traced to dye from the hospital’s brandmarks on a batch of new cotton nappies (which had been washed before being marked)”.1 The dye was understood to have been “absorbed into the infants’ circulation via their raw umbilical areas”. As it turns out, nappy dyes have been incriminated in MetHgb.4 The POISINDEX® System (Thomson Micromedex, Denver, Colo, USA) revealed that the aniline group is among the compounds capable of causing MetHgb through any portal of entry, even intact skin (Stephen Gibbins, Poisons Information Specialist, Victorian Poisons Information Centre, Melbourne, personal communication). Another textile ink, aminophenol, may also lead to the disorder.5 Other common causes of MetHgb are given in the Box. Causes of methaemoglobinaemia (MetHgb) MetHgb can be genetic, through faulty haemoglobins or cytochrome deficiency. On exposure to appropriate exogenous toxins, heterozygotes tend to be particularly susceptible, as are infants per se without any such trait.4 Nappy dyes have been incriminated.4 In the United Kingdom between 1961 and 1980, the most common industrial causes of acquired MetHgb were chloroaniline, p-toluidine, nitrobenzene, nitrochlorobenzene, nitrates and amines — frequently by dermal exposure. Nitrates, nitrites in food, local anaesthetics (particularly benzocaine), and other medicinal agents, such as chloroquine, dapsone, para-aminosalicylic acid and resorcinol, have also been implicated.5,6 Aniline dyes were discovered in the 19th century and found useful for inks and for dyeing leather, which they penetrate as they do living skin. They were also used in wood stains, textiles, and oriental rug-making. Because aniline dyes are not indelible and are toxic, they are being superseded by chrome dyes. The hospital linen room concerned1 reports that they now apply dry, colour-fast pigments using heat-pressed or stamped “transfers” (Administrative Officer, Metropolitan Linen Services, Brisbane, personal communication). It is not possible to confirm what the “Baby blues”1 dye was — aniline or other. Nor can we now establish whether the dusky colour of the babes was due to absorbed dye circulating passively. Circumstantially however, MetHgb would seem to have been a likely possibility.4 In differential diagnosis, when cyanotic infants or adults without cardiorespiratory signs are encountered, we should not forget MetHgb, drugs and chemical toxins. In patients with MetHgb, the blood will be “chocolate-brown”; and yes — do remember to take a sniff for that exhaled acetone.2

Ivan Cher

To the Editor: We recently read with concern the article by Scott and Ward on public reporting of hospital outcomes.1 While we do not want to enter into the debate about whether the public release of hospital performance reports is beneficial or harmful, we would like to address some issues relating to the accuracy of administrative data. The authors stated that “data are often [our italics] inaccurate, incomplete, or provide insufficient clinical detail” and that the “accuracy of diagnosis coding is vari-able”. They also mention the potential for “gaming” or “up-coding” by hospitals to make their institutions look better in public reports. We believe the authors’ argument regarding coding inaccuracies is flawed. One of the articles they cited was not about coding accuracy but about mortality differentials between metropolitan and non-metropolitan regions.2 Another cited article quoted an example of a 100% miss rate for coding of dementia as a comorbidity that was based on only three cases.3 As the authors later stress in their article, it is important that sample size be considered when interpreting data, to ensure that the effects of random error are minimised. We agree with the authors’ final point that distinguishing complications from presenting diagnoses (or comorbidities) is currently difficult using hospital coded data. However, the Victorian and Queensland hospital data collections now include an “alpha flag”, which is a letter attached to each coded diagnosis to indicate whether the diagnosis was present on admission to hospital or whether it arose during the episode of care. Moves are underway to introduce a minimum national requirement to use the alpha flag as one method of identifying complications arising from medical or surgical care. There are a number of national and state initiatives that aim to ensure the national morbidity data collection is as accurate as possible and that it provides data directly related to its purpose (and therefore not necessarily useful for other purposes). There is currently a national debate about the purpose of this collection. It is clear that there are issues surrounding the capture and coding of hospital data that are not well understood by data users. Because of that misunderstanding, reports such as the one by Scott and Ward paint an unjustifiably bleak picture of the quality of the data.

Kerry Innes · Kirsten McKenzie · Sue Walker

In reply: Innes and colleagues accuse us of overstating the potential inaccuracy of coded administrative data. They refer to state and national initiatives underway to ensure such accuracy, but offer no hard statistics that would reassure us that such data, in their current form, are as accurate as they need to be for purposes of quality monitoring and public disclosure. Until they do, we feel we have good reason to recommend caution in light of the few published Australian reports that are available (which we cited1,2), together with other research3 and feedback from clinical directors, about significant error rates when coded diagnoses are audited by clinicians or compared with independent datasets maintained by clinicians (Professor David Johnson, Director of Nephrology, and Dr Paul Garrahy, Director of Cardiology, Princess Alexandra Hospital, personal communication). In Queensland, formal regular audits on coding accuracy were initiated only in October 2005. They involve small numbers of randomly selected charts from each hospital and focus on specific coding issues identified for each hospital (Professor Stephen Duckett, Executive Director of Reform and Development, Queensland Health, personal communication). While we welcome (and were aware of) the introduction of “alpha flags” to distinguish in-hospital complications from pre-existing conditions, these remain a recent development (especially in Queensland), and others with considerable experience in their use express caution in interpreting results in the absence of rigorous validation.4,5

Ian A Scott · Michael Ward

To the Editor: Your 3 July issue featured task substitution and task transfer. It is a principle of commercial organisation that if a task can be standardised, it can be delegated, automated or computerised, provided there is good central management, supervision and communication. However, I suspect that this is only a part of the new face of general practice, as we are being expected to undertake tasks for which my age cohort (I am 54) was neither trained nor prepared. Practice administration is far more complex than ever before, and most of the clinical caseload has shifted from episodic care of infections and surgical conditions to the long-term systematic management of chronic cardiovascular, respiratory, musculoskeletal, endocrine, and other illnesses. Diabetes is a good example. Managing patients with diabetes requires high-level skills in practice management and protocol-driven chronic disease care. I suggest that a basic and fundamental difference between doctors and allied health personnel is that allied health personnel are extremely comfortable with protocol-driven chronic management while doctors, especially of my age cohort, are more focused on detecting and dealing with difference, variation and abnormality in our patients’ health. Rather than force all groups to do the same tasks, is it not better to build on their skills and interests in a logical and structured manner? General practices are no longer only places where general practitioners work; they are evolving into teams of GPs, other doctors, administrators, allied health personnel and office and information techno-logy staff, who work together in an integrated and coordinated way for the benefit of patients. This provides mutual support, flexibility of work hours and increased job satisfaction for all. Our practice has been steadily working towards this for the past 20 years.

Christopher D Hogan

To the Editor: A young man, a minor when sentenced in Sydney, was diagnosed with lymphoma soon after incarceration. Appropriate treatment was initiated, including pretreatment collection and storage of his semen. A local newspaper report that his sperm was collected and stored at taxpayers’ expense prompted outrage in some sections of the community. In response, the New South Wales Government drafted the Corrective Services Legislation Amendment Bill 2006, which would make it a crime for an individual imprisoned or awaiting sentencing for a “serious indictable offence”, such as homicide, rape or terrorism to store “reproductive material” (semen or ova).1 It is routine (many would say mandatory) for men of reproductive age who are about to undergo therapy for cancer to be offered the option to store semen. Without this option, male cancer survivors might be unable to father their own offspring. There is no current routine technology for storing unfertilised ova. In current practice for male prisoners, semen is stored before commencing treatment for cancers or similar conditions that may induce temporary or permanent infertility. This is the accepted standard of care, offered before such treatment to men who may not have completed their families. It is not current practice in NSW to store prisoners’ semen in any other circumstances. The proposed Bill will discriminate against prisoners in the quality and costs of their health care. Members of our community who require chemotherapy for cancer are offered collection and storage of their semen, provided free of charge by several public services in NSW. Under the proposed Bill, prisoners are required to pay storage fees during their imprisonment, even if their sperm were placed in storage before their incarceration. Discriminating against certain prisoners by demanding payment for otherwise free services could be seen as a “cruel and unusual punishment”. The NSW Legislative Assembly passed the Bill on 25 May 2006. Medical, legal and human rights organisations, and individuals expressed concern to parliamentarians. In the Legislative Council on 7 June 2006, a majority vote referred the Bill to the General Purposes Standing Committee No. 3. This Committee has received submissions and will provide recommendations as to how the Bill should proceed. If passed into law, the Bill would breach the principle of equivalence of health care for prisoners. The Australian Medical Association position statement on the Health care of prisoners and detainees states: “The duty of medical practitioners to treat all patients professionally with respect for their human dignity and privacy applies equally to the care of those detained in prison, whether convicted or on remand, irrespective of the reason for their incarceration.”2 I argue that the Bill implies an intention to rid society of “criminal seed” and begins a move towards eugenics. If our society really accepts the idea that inmates of correctional facilities may one day return to a full and productive life, then it is unreasonable to deny them the possibility of having their own children because they developed a serious cancer. If this legislation is passed, a discriminatory practice of medicine according to convict status will be enshrined in NSW law.

John E J Rasko

To the Editor: Quality data about patients with injuries requiring hospitalisation is vital to injury policy and prevention strategies.1 The ICD-10-AM is used in Australia to assign codes to diagnoses, procedures, and causes of injury recorded in patient medical records.2 This coded hospital morbidity data provides a key surveillance tool for injury researchers. ICD classifications are designed for statistical reporting and are required for classifying all information encountered in hospital medical records. When insufficient information is available in the medical record to assign specific codes, the use of residual “Unspecified” categories helps to achieve this. Detailed and accurate documentation provided by clinicians in patient medical records is imperative to produce high quality coded data.3 Poor documentation in medical records has been shown to decrease data quality by contributing to an overuse of “Unspecified” codes.4 This is especially so for documenting external cause of injury, which may not be seen as critical to the patient’s care by the treating clinician, with the result that the relevant detail is incomplete or omitted altogether. We aimed to identify the level of precision of coded injury data in Australian hospitals. Using the 2003–2004 national morbidity dataset, 445 098 records containing an injury and an external cause classified by intent were found (Box). At a broad intent level, the majority of injuries were assigned to a specific mechanism code, although in two intent categories, “Accident” and “Assault”, 11% and 13% of injuries, respectively, were assigned to “Unspecified” categories. It is concerning that 45 297 of the injuries requiring hospitalisation lacked adequate documentation in the medical record to permit meaningful code assignment for cause of injury. A significant lack of precision was evident in recording mechanisms of accidental falls and poisonings (across all intents). A quarter of falls and 20% of poisoning cases had no specific information about the causal mechanism or substance. Being the most commonly reported accident mechanism, falls of unspecified cause represented 11% of accidents overall. This lack of detail is particularly concerning given the significant national priority now placed on falls and poisoning injury prevention.5 It is essential that clinicians and coders alike are aware of documentation and coding problems related to capturing data on cause of injury. By working together to improve the quality of injury-related coded data (through improved clinical documentation), accurate and comprehensive information pertaining to the circumstances surrounding injury events requiring hospitalisation will benefit injury policy and prevention initiatives. Precision of recorded cause of injury data across selected ICD-10-AM categories2 for 2003–04 ICD-10-AM categories Specified Unspecified Total Intent Accident 347 781 (89.2%) 42 078 (10.8%) 389 859 Intentional self-harm 29 018 (99.6%) 130 (0.4%) 29 148 Assault 19 385 (87.2%) 2 841 (12.8%) 22 226 Undetermined intent 3 617 (93.6%) 248 (6.4%) 3 865 Total 399 801 (89.8%) 45 297 (10.2%) 445 098 Mechanism Accidental falls 130 089 (74.6%) 44 336 (25.4%) 174 425 Poisoning (all intents) 31 111 (80.0%) 7 798 (20.0%) 38 909

Kirsten McKenzie · Leith F Harding · Susan M Walker · James E Harrison · Emma L Enraght-Moony · Garry S Waller

To the Editor: The recommendation for managing acute ST-segment-elevation myocardial infarction with percutaneous coronary intervention (PCI) is that the door-to-balloon inflation time should be 90 minutes. However, it can be up to 120 minutes, depending on when patients present to the emergency department (ED) after the onset of their symptoms.1 In such cases, an alternative immediate reperfusion strategy — fibrinolysis — should be considered. At first glance, a door-to-balloon time of 90 minutes seems readily achievable, but what if the patient presents after hours, or presents to a hospital without PCI facilities? The time required to refer the patient for PCI, organise ambulance transport and call in cardiac catheterisation laboratory staff can be considerable. In the PRAGUE-2 trial from the Czech Republic, the average door-to-balloon time was 97 minutes.2 The DANAMI-2 study from Denmark had a cohort of 27 080 patients and had door-to-balloon times of about 114 minutes for those patients transferred to another facility.3 The National Registry of Myocardial Infarction 4 investigators reported a median door-to-balloon time of 185 minutes for American patients transferred to centres capable of PCI, and a door-to-balloon time of less than 90 minutes for only 3% of patients.4 Doctors working in EDs without onsite access to PCI need to know the door-to-balloon times of the institutions to which they refer patients for PCI. Centres performing PCI may not be forthcoming with this information, as they have a vested interest in keeping their numbers up for PCI. Alternatively, this information may not be known to the clinician accepting the patient for PCI. In addition, doctors in EDs who opt to transfer their patients have the burden of organising transport for potentially unstable patients who may develop lethal arrythmias. As door-to-needle time for thrombolysis has become a clinical indicator for EDs, perhaps door-to-balloon times can be a clinical indicator for cardiac catheterisation laboratories. Finally, it is important to note that in some patients requiring urgent coronary artery reperfusion, the first electrocardiogram (ECG) is not diagnostic, so a more pragmatic indicator would be diagnostic ECG-to-balloon time. This would require an enforcement of the current recommendations for an ECG to be performed and critically reviewed shortly after a patient presents with symptoms suggestive of an acute coronary syndrome.

Jayantha I Weeraratne

In reply: We agree with the points highlighted by Weeraratne, and these have been broadly addressed within the new National Heart Foundation of Australia and Cardiac Society of Australia and New Zealand Guidelines for the management of acute coronary syndromes 2006.1 The guidelines emphasise the need for appropriate systems of care which are regionally based, have formal links with specialist centres, include appropriate monitoring, feedback and quality improvement components, and are sensitive to the cultural and personal beliefs and wishes of individual patients. Clinicians do need to know the achievable door-to-balloon times for primary percutaneous coronary intervention within their local contexts, and if there is any doubt about the timely availability of this treatment for patients with ST-segment-elevation myocardial infarction, the guidelines recommend that fibrinolysis be given promptly.

Philip Aylward · Constantine N Aroney

To the Editor: The editorial by Coiera and Westbrook raises some important points.1 Appropriate models of governance (vis-à-vis regulation) surrounding clinical software are required if we are to drive innovative technology on a course that is safe and effective for patients. The success or failure of an information system depends on the organisational context in which it is placed.2 The people, the work processes and the technology must be viewed as integrated elements of one system that aims to improve health quality and, importantly, do no harm. In essence, each element provides an additional layer of quality control and safety, and these work together to avoid adverse events. By law, we require medical practitioners to be registered. By law, we require health facilities to be accredited and licensed. We should also, by law, acknowledge clinical software as being part of the jigsaw puzzle and require that it too be regulated. Because of the complexity of the task, in-house development of clinical systems is unattractive to organisations, leaving vendor solutions as the alternative.3 Currently, there is no apparent active engagement between software developers, government, clinicians and funding bodies to establish a transparent and sustainable program of decision-support software development in Australia.4 Beilby et al recommended a generic standards-based “middleware” that sits outside all clinical desktop software systems and supports the exchange of information with other clinical systems and knowledge repositories.4 This would be the gold standard. Inextricably linked with this would be a regulatory framework to compel software suppliers to comply with the standard. Without this, health care organisations are vulnerable to the whim of software vendors, each with different standards, capabilities and knowledge capital. Decision-support tools to supplement memory and record clinical information and results can help standardise clinical care and reduce human error by ensuring that uniform, evidence-based practices are adopted.5 Electronic decision-support systems are currently espoused as one of the keys to good quality and safe health care.4 The health system needs this innovative technology. However, if the health system is to avoid duplication, fragmentation and inconsistencies associated with multiple standards for electronic decision-support systems and other clinical software, then we must advocate for workable standards and legislative frameworks on which to build the technical solutions. We owe this to the health professionals using these systems, who may otherwise make a wrong turn, and we owe it to the patients at the end of the line.

Karen L Fox BAppSc(HIM)

To the Editor: I wholeheartedly agree with Giles at al1 regarding the need for universal HIV antenatal screening in Australia. However, it is worth noting that, at least in private practice, there is already a significant amount of antenatal HIV screening taking place. In November 2005, several Medicare Benefits Schedule (MBS) item numbers were introduced for antenatal screening for infectious diseases including HIV testing. I am unaware of any specific HIV-testing restrictions relating to national policy (other than the obligations of informed consent and such like) attached to these item numbers. Four MBS item numbers for “microbiological serology during a pregnancy” may include HIV testing (69405, 69408, 69411 and 69413), and one MBS item number (69415) must include HIV testing. From November 2005 to June 2006, there were 137 732 claims for antenatal serological tests, of which at least 46 085 (33%) included HIV testing (see Box).2 There is some variation from state to state (New South Wales, 25%; Victoria, 34%; and Queensland, 38.5%). Significant state-to-state variation of claims for different Medicare items is not unusual but, in this case, it does not appear to follow any pattern (of the epidemiology of HIV infection in Australia). I suspect that the proportion of pregnant women having HIV tests is closer to 50%, assuming that at least some of the other item numbers claimed included testing for HIV. It would be worthwhile taking these figures into account when formulating national policy. Medicare items for antenatal serological testing, which may include HIV testing — number of items processed in Australia from November 2005 to June 2006 by state Item NSW VIC QLD SA WA TAS ACT NT All states 69405 2 664 1 905 1 291 149 919 177 142 269 7 516 69408 2 375 1 191 1 080 115 769 205 170 114 6 019 69411 16 169 5 493 6 758 549 1 890 913 912 284 32 968 69413 15 874 10 267 9 488 2 792 4 171 1 229 588 735 45 144 69415 12 283 10 138 11 695 3 095 6 710 784 424 956 46 085 Total 49 365 28 994 30 312 6 700 14 459 3 308 2 236 2 358 137 732 Item 69415 must include HIV testing.

Len D Moaven

To the Editor: We report three cases of locally acquired Entamoeba histolytica infection in men who have sex with men (MSM) in Sydney, New South Wales. E. histolytica is an invasive pathogenic amoeba that can cause invasive intestinal and extraintestinal amoebiasis. Entamoeba dispar is morphologically identical but is considered non-pathogenic and non-invasive.1 The three patients presented with a 1–3-week history of diarrhoea and abdominal pain. Routine bacterial cultures were negative for pathogens. Ova, cyst and parasite investigations showed cysts and trophozoites of E. histolytica/dispar complex in permanently stained, fixed faecal smears. Stool samples were tested for E. histolytica and E. dispar by polymerase chain reaction (PCR), using a previously described method.2 All three patients were positive for E. histolytica by PCR; sequencing of the amplicons verified the presence of E. histolytica DNA. The three patients presented within a 12-month period in 2005–2006. All were homosexually active men (ages, 31–53 years) who lived in inner Sydney. None had a history of overseas travel within the previous 5 years, suggesting that the infections were locally acquired. High rates of intestinal parasitism are found in MSM throughout the world. Oral–anal and oral–genital sexual practices are reported to predispose to infection with enteric pathogens, particularly protozoa. A 1991 study reported a higher prevalence (37%) of E. histolytica/dispar complex in a homosexual population in Sydney when compared to non-MSM.3 However, that study did not differentiate between the two species E. histolytica and E. dispar. Amoebiasis has become endemic in MSM in Japan and causes significant morbidity and mortality; complications such as colitis and liver abscesses occur more frequently in homosexual and bisexual men than in heterosexual men.4 Similar findings on amoebiasis are reported from Taiwan, with MSM at increased risk for invasive amoebiasis and intestinal colonisation with E. histolytica.5 The discovery of E. histolytica infection in MSM in Australia is of public health concern and highlights the importance of continued surveillance, as the organism has the potential to become endemic in the gay population and to cause significant morbidity. Clinicians should also be aware that E. histolytica is present in urban settings in Australia and should be included in differential diagnoses.

Damien J Stark · Rashmi Fotedar · John T Ellis · John L Harkness

To the Editor: Meningitis is one of the most serious infections in young children. The annual incidence of Haemophilus influenzae type b (Hib) meningitis between 1984 and 1988 was 150 per 100 000 population in Aboriginal children and 27 per 100 000 in non-Aboriginal children younger than 5 years.1 A conjugate Hib vaccination program was introduced in Western Australia in January 1993, before a nationwide program commenced in July 1993. Subsequent marked declines in incidence of Hib meningitis have been reported.2-4 However, there are no recent reports on trends in overall admissions for meningitis. The WA Data Linkage System (WADLS) encompasses statewide population-based record linkage of the statutory birth and death registers, midwives’ notification system, and hospital morbidity database,5 and is one of few such resources worldwide. As part of a larger study to determine the burden of infection in a cohort of births between 1990 and 2000 using the WADLS, we investigated hospitalisation for all-cause meningitis (International classification of diseases, 9th revision, diagnosis codes 003.21, 036.0, 047, 049.0, 054.72, 320-322) in 17 296 Aboriginal and 252 775 non-Aboriginal children younger than 2 years between 1992 and 2000. In Aboriginal infants (< 12 months), the meningitis rate fell by 41% between 1992 and 1993–1994 and by a further 54% in 1995–1996, and has remained stable since (Box). In Aboriginal children aged 12–23 months, rates declined by 44% between 1993–1994 and 1995–1996 and again by 50% in 1997–1998, and no meningitis admissions were reported in 1999–2000. In non-Aboriginal infants, meningitis rates declined by 36%, from 1.8 per 1000 child-years in 1992 to 1.2 per 1000 child-years in 1993–1994, with a further 50% decline in 1997–1998, since when rates have remained stable. Rates declined by 57% between 1992 and 1993–1994 in non-Aboriginal children aged 12–23 months, declined a further 47% in 1995–1996, and have since remained stable at about 0.2 per 1000 child-years. With the decline in meningitis admissions, the disparity between Aboriginal and non-Aboriginal children has narrowed: the relative rate (RR) of Aboriginal to non-Aboriginal meningitis admissions fell from 7.3 in 1992 to 5.0 in 1999–2000 in infants, while in children aged 12–23 months, the RR was > 7.0 in 1993–1996, fell to 3.0 in 1997–1998, and was indefinable in 1999–2000 (Box). In the absence of other relevant interventions, we attribute declines in meningitis admissions to the introduction of Hib vaccine. This is supported by other studies showing a reduction in Hib meningitis following vaccination.2-4 Retrospective data provide an opportunity to assess overall trends in admissions. Future linkages with immunisation and laboratory data will allow us to investigate pathogen-specific admissions and evaluate vaccination programs. Our findings show that substantial improvements can be achieved given government commitment to implement appropriate preventive measures. Adequate funding and continued commitment is needed to ensure these measures are accessible to all WA children. Hospital admission rate for meningitis in Aboriginal and non-Aboriginal children aged (a) < 12 months and (b) 12–23 months in Western Australia, 1992–2000 Relative rate of Aboriginal to non-Aboriginal admissions is shown at the top of each graph.

Hannah C Moore · Deborah Lehmann

To the Editor: The current rate of caesarean sections in Australia (29% of all live births) is higher than the rate in other similarly affluent countries.1 In addition, the rate is continuing to increase; for example, it was less than 20% in 1993.1 Some commentators have suggested that this increase is partly a result of caesarean sections undertaken for non-medical reasons, such as patient demand.2,3 We examined trends in the rates of caesarean section for low-risk women using population-based perinatal data for Queensland over 15 years between 1990 and 2004. Our aim was to assess whether caesarean sections were becoming more common among women with no obvious medical indication for the procedure. The increase in caesarean sections among low-risk women was most dramatic in the private health care sector, where the percentage increased from 10% to 19% (Box). This represents an average annual increase of 4.6% (95% CI, 4.3%–5.0%). In the public health care sector, the increase was less — from 6% to 8% — an average annual increase of 2.4% (95% CI, 2.0%–2.7%). The increase in the private sector in Queensland was similar to the increase reported in the United States.4 The appropriate use of caesarean section, as for any medical intervention, should be based on evidence about the benefits and harm, with doctors, women and their families choosing a method of delivery after considering balanced information on potential outcomes of each method. There is continuing debate about the feasibility of randomised trials to clarify the benefits and harm of caesarean deliveries among low-risk women.2 Opposition to such trials is based mainly on ethical concerns about inflicting a surgical procedure on healthy women based only on randomisation. Non-randomised studies have compared outcomes of caesarean section versus vaginal delivery. However, their results are inconclusive because of the difficulty of distinguishing the effects of factors that influence the selection of delivery method from the effects of the delivery method itself (confounding by indication).3,5 In the absence of randomised trials, non-randomised studies that remove this potential bias by restricting the sample to women who remain at low risk throughout the pregnancy and delivery, according to clearly defined criteria, may provide useful information. They would need to assess both short-term and long-term outcomes. Until such better evidence is available, it is impossible to judge whether or not the current increase in caesarean section rates among low-risk women is desirable. Caesarean section rates among low-risk* women in Queensland, 1990–2004 * Low-risk births were defined as singleton, full-term (37–40 weeks’ gestation), vertex delivery with no reported medical risk factors or complications of labour or delivery, based on a list compiled by Declercq and colleagues.4 Women who had a previous caesarean delivery were excluded from the low-risk group.

Trisha C Johnston · Michael D Coory

To the Editor: The incidence of obesity is rising, and physicians are likely to face the problem of obesity-related glomerulopathy (ORG) recently illustrated by Tran.1 But how can the clinician distinguish ORG from primary (idiopathic) focal segmental glomerulosclerosis (FSGS)? Both may present with nephrotic-range proteinuria, but the prognosis and choice of treatment may differ. To date, the largest published study comparing ORG with primary FSGS is one by Kambham et al.2 In an analysis of 6818 renal biopsies, 71 patients with ORG were identified and compared with a control group of 50 patients with classic FSGS. The study showed that ORG less frequently progressed to end-stage kidney failure, with a 5-year renal survival rate of almost 90% (compared with about 50% in primary FSGS).2 While weight loss can reduce hyperfiltration and albuminuria in ORG,3 spontaneous remission is uncommon in primary FSGS.4,5 Does every obese patient with nephrotic-range proteinuria have ORG and an “indolent” course? The degree of weight loss reported in the case described by Tran may not be achievable or sustainable in most obese patients. Do we have the luxury of waiting to assess the impact of weight loss on proteinuria? In about 50% of patients with primary FSGS, the serum creatinine level doubles after an average of 39 months.2 Furthermore, patients with primary FSGS and nephrotic-range proteinuria who do not achieve remission have a 5-year renal survival of only 50%, compared with almost 100% for those who attain remission.4 In addition, patients treated with corticosteroids (with or without cyclosporin or cyclophosphamide) have higher remission rates (30%–63%) than untreated patients (11%–14%).4,5 Therefore, a delay in introduction of specific therapy is not ideal. There are some clinicopathological differences between ORG and primary FSGS that may help distinguish the two entities (Box). However, Kambham et al found that only two parameters were independently significant: serum albumin level and age.2 Although their study was based on a US population, it serves to demonstrate the principle that the major distinguishing feature between ORG and primary FSGS is the presence of full-blown nephrotic syndrome in primary FSGS (as demonstrated by the severity of hypoalbuminaemia). Obese patients have a similar risk of developing primary FSGS to people in the general population, and patients with nephrotic syndrome (particularly older adults) should not be presumed to have ORG and treated with weight loss alone. Certain pathological findings in a renal biopsy are helpful, but not definitive, in distinguishing ORG from primary FSGS. A biopsy would also exclude other treatable causes, such as minimal change disease. In addition to treatment with angiotensin-converting enzyme inhibitors, immunotherapy should be considered for obese, nephrotic patients, after discussing the potential risks and benefits with a nephrologist. Clinicopathological differences between ORG and primary FSGS* Parameter ORG Primary FSGS Mean age at presentation (years)† 42.9 32.6 Ethnicity White (%) 74 52 African American (%) 21 22 Nephrotic syndrome (%) 5.4 54 Mean 24-hour protein excretion (g) 4.1 6.9 Mean serum albumin level (g/L)† 39 29 Mean serum cholesterol level (mmol/L) 5.9 8.6 Presence of pedal oedema (%) 35 68 Mean degree of segmental sclerosis (%) 10 39 Proportion of cases with glomerulomegaly (%) 100 10 Mean arteriosclerosis score (range, 0–3) 1.34 0.98 Mean degree of glomerular podocyte foot process fusion (%) 40 75 ORG = obesity-related glomerulopathy. FSGS = focal segmental glomerulosclerosis. * Adapted from Kambham et al.2 † Independently significant.

Andy K H Lim

To the Editor: Recent articles in the Journal by Kurrle1 and Finn and colleagues2 referred to advance care directives aiding the management of acute illness in elderly residents of aged care facilities. It is our experience that these directives are often unhelpful in elderly patients and, outside certain progressive medical conditions, can result in triage of elderly patients to inappropriate lower levels of care. In chronic medical conditions where the clinical course allows time for patient or family understanding, and the course of organ failure is predictable, then certain supportive but ultimately futile therapies can be avoided by instituting an advance care directive that specifically excludes them. However, these directives are less helpful in acute illnesses. They usually refer to “intensive care”, and “life support”, sometimes specified as mechanical ventilation, dialysis or cardiopulmonary resuscitation. These “general” advance care directives fail, as they assume that prognosis is immediately apparent, and that treatment is “all or nothing”, both of which assumptions are clearly untrue. Determining an accurate prognosis for recovery from a critical illness is difficult and takes time. It involves diagnosing the cause of the illness, quantifying the severity of comorbidities and, most importantly, assessing response to initial treatment. Whether severe sepsis is arising from the urinary tract or abdominal cavity may not be apparent initially. Many elderly patients survive severe septic shock caused by urosepsis with haemodynamic monitoring and short-term high-dose vasopressors. It is also not possible to distinguish which patients with severe respiratory failure will respond to non-invasive ventilation. We followed up critical care patients aged 75 years and over who survived to hospital discharge over a 12-month period and confirmed that acceptance of critical care admission in elderly people is high (unpublished study; details available from the authors). This is the very population that, in our experience, frequently says they do not want to be placed on “life support”, if asked when well. Together with the fact that an accurate prognosis takes time, then a prudent approach should begin with the presumption of aggressive treatment for acutely unwell elderly patients, rather than a presumption of limited therapy or palliation. Advance care directives that refer to therapies need to be specific and to recognise that critical care therapy can be graduated and readily terminated once a more accurate prognosis is known. Furthermore, some critical care therapies, such as non-invasive ventilation and high-concentration oxygen, can significantly improve patient comfort while management plans are formulated. In our experience, patients and their families are often very surprised when they understand the full implications of an advance care directive that refers to generic therapies, such as cardiopulmonary resuscitation and “intensive care”.

Andrew W Holt · Alnis E Vedig

To the Editor: I note with interest Australian medical students’ difficulties in gaining access to patients, as documented by several authors in recent months.1-3 There is an alternative explanation for this paucity of access, and that is the culture in Australian teaching hospitals. I graduated from the University of Otago Dunedin Medical School in New Zealand 9 years ago, and had a somewhat different experience. The hospitals attached to the Dunedin Medical School were the equivalent of one medium-sized acute hospital in Australia, one rehabilitation hospital, and a small peripheral regional hospital. These facilities taught up to 200 clinical medical students — a high student-to-patient ratio. However, we did not face the barriers that Australian students face in gaining access to patients, and so still had a world-class medical education. The system in Otago differed from that in Australia in several ways. If a patient was having an investigation, we accompanied them. Likewise, if they were seeing a staff member, we would often stay. There was a culture where every patient admitted to hospital expected to see a medical student. We approached the patients directly to ask permission to see them, rather than being turned away by nursing staff. Consequently, in non-obstetric patients, I encountered only one refusal to see me in my clinical years, and this was after I had started taking a history and asked about tranquilliser use in too much detail! We also did not wait to see only patients who were ideal teaching cases — patients who speak English, do not have dementia, are not unwell, are not busy, and who have a particularly interesting condition are rare anywhere. If a patient had dementia or was unwell we familiarised ourselves with their history, then saw the patient over the course of several days. If there were visitors, we asked about an appropriate time to come back. We saw all routine cases, as these reflect the real workload of doctors. By comparison, in my years as a resident and registrar in Australia, I have seen very few medical students, and have seen many teaching and learning opportunities pass by. In conclusion, we need to examine the role of medical students closely, and look at ways to facilitate their access to existing patients. There is something to be learned from all patients admitted to hospital. The culture of teaching hospitals and the expectations of students, staff and patients should reflect this fact.

Sarah J Abrahamson

To the Editor: The 2005 Cochrane review Home-like versus conventional institutional settings for birth1 has been cited in the public media to claim that birth centres are less safe than labour wards as there was an increased risk of a baby dying during or immediately after childbirth.2 This “headline-grabbing” statement is false. Firstly, this finding from the systematic review did not reach statistical significance.1 Secondly, the outcomes reviewed were related to the allocated place of birth, not the care provided. This fact is critically important, as 48% of women who were booked to have their baby in a birth centre did not give birth there.1 This is a predictable effect of the intention-to-treat principle. However, such high rates of “treatment contamination” negatively affect confidence in the study results.3 Additionally, the vast majority of baby deaths examined in the Cochrane review happened before labour and thus had nothing to do with care during childbirth. One might wonder whether there was a real increased perinatal mortality rate resulting from delayed transfers from birth centres.1 The analysis found 41 deaths in total, but only six that occurred in normally formed babies who reached term (these are the only babies who are eligible to be born in a birth centre). Three of these deaths were associated with birth centre care, and three with standard labour care. The interpretation of this Cochrane review raises questions about the validity of the underlying randomised controlled trials. In this experimental design, researcher control should ensure that people receive the specific treatment that was planned for them (treatment fidelity).4 The Cochrane handbook gives no guidance as to how to evaluate either the quality of the researchers’ definition of the planned treatments, or the fidelity between the treatments provided and the researchers’ plan.3 Most of the trials that formed the basis of the Cochrane review did not adequately define their treatments, nor adequately control the treatments provided to either group. It is not clear how the birth centre trials could sensibly be considered to have been scientifically controlled. The reviewers attempted to deal with this critical point by claiming that they were looking only at the effect of the “setting”, but their question clearly states that they were examining the effect of “care within a setting”.1 We conclude that the Cochrane review of the setting for birth is unreliable because of the weaknesses of the underlying trials. Rather than using questionable research to attack birth centres, it would be more constructive to engage in rigorously designed research that could provide robust evidence on the safety of all forms of maternity care, including standard medical care.

Kathleen M Fahy · Sally Tracy

In reply: Fahy and Tracy highlight the lack of high-level evidence about the relative safety of different models of maternity care. But in criticising the Cochrane review, it is important not to “shoot the messenger”. There is no doubt that the Cochrane review is not ideal but, like it or not, it remains the best evidence we have. The review of 8677 women in six randomised trials found a relative risk (RR) of perinatal death of 1.83 (95% CI, 0.99–3.38) in birth centres versus conventional institutional settings. In the 3332 pregnancies assigned to continuity of care by midwives who did not also work in conventional delivery suites, the RR was 2.38 (95% CI, 1.05–5.41).1 It would be fair to say that such findings should lead to real concerns about lack of safety rather than reassure the unbiased observer. Other published evidence has raised similar concerns. A retrospective review of over 183 000 low-risk births in Stockholm, Sweden, found a statistically significant fourfold increase in intrapartum fetal mortality in women planning birth centre care compared with those planning standard care (three intrapartum deaths in 3256 babies of women planning birth centre care versus 36 deaths in 180 380 babies of those planning standard care).2 The increase in intrapartum mortality was almost sevenfold for primigravidae. These findings led to evidence-based changes in the organisation of the birth centre involved to minimise the identified risks. To paraphrase Fahy and Tracy, rather than criticising the best available evidence reviewing birth centre outcomes, it would be more constructive to engage in rigorously designed research to assess how risk might be minimised in all forms of maternity care.

Andrew F Pesce

To the Editor: I was very concerned to read the letter from Mazza et al1 regarding “positive” family planning and feel I must make a comment. The authors are well known for their work in the area of women’s sexual and reproductive health, but I would like to challenge some of the points they have made. The first point: whether intervention by general practitioners would be appreciated by younger women not yet interested in motherhood. I believe it is part of the role of doctors to inform, even when a person may not be ready for the information. Telling 20-a-day smokers that they are not doing their body any favours doesn’t go down well with some people, but even this brief intervention can change behaviour and save lives. The second point: whether GPs can respect patients’ autonomy. Every day I speak to women who have been offended and upset by doctors who have said something awkwardly, or imposed their own values, or been downright offensive. That won’t change, and recommending that well informed and tactful doctors wait before imparting vital information until the rest of the world lifts its game means we will all be waiting a long time. Part of the art of medicine is judging the audience and knowing the perfect point in a consultation to speak, and how to say it. What can be more appropriate, when seeing a woman in her late 20s who has requested a repeat prescription for the contraceptive pill, than to ask casually (as one is unrolling the sphygmomanometer cloth), “So, do you think there might be any children in your future?” The usual response, as detailed in Cannold’s book,2 is an emphatic “yes”. The next question, “Have you got a time scale when you would be looking at that?”, may give the opportunity to mention such things as rubella vaccination, smoking and folate supplements. And if the woman indicates that pregnancy would be on her to-do list at age 38, then a reasonable and non-harassing response could be, “Could we talk about fertility rates at that age?” The third point: doctors reinforcing the dominant paradigm (of years ago) of the woman as childbearing machine, by asking about a woman’s intentions. This seems to me as misguided as not asking about suicidal ideation in case we make it happen. By all means do research, but don’t ask doctors to be silent about this important issue until the sociologists have spent another 10 years on it. By that time, it will be too late for a lot more women.

Angela M Cooney

To the Editor: Since my article on medical abortion was published in the Journal,1 I have received some information from colleagues, and from women who have undergone abortions, about the current practice of medical abortion in Australia. I believe this may be of interest to readers of the MJA. More than a dozen practitioners have informed me that they have used misoprostol or methotrexate/misoprostol combinations to induce early abortion (before 9 weeks’ gestation) outside of hospitals, and a number of women have reported undergoing such abortions. The number of cases involved in this anecdotal sample is at least several hundred annually. Induced abortion using methotrexate/misoprostol was practised in the United States up until the introduction there of mifepristone/misoprostol regimens in 2000.2,3 There is wide experience of this drug combination reported in the medical literature, and the consensus is that, in the short term at least, it is safe and effective, although less effective than the mifepristone/misoprostol combination.2-4 Both drugs are licensed for purposes other than abortion in Australia, as elsewhere (misoprostol for treatment of gastric ulceration; methotrexate as a cytotoxic agent, and for psoriasis and rheumatoid arthritis), but the use of drugs “off-label” is an acknowledged medical practice.5 Misoprostol in particular is widely used in obstetrics for cervical ripening and treatment of postpartum haemorrhage.5 These early abortions take place “under the radar” in the sense that there is no specific Medicare item number; the administration of the drugs occurs in the course of a standard consultation. There is nothing irregular about this, but it does mean that the inaccurate data we currently have on the number of abortions performed in Australia are even more inaccurate than initially thought. It should be noted also that misoprostol alone or in combination with other prostaglandins is being used in Australian hospitals, as overseas, in a more evident manner for the induction of late abortion in cases of severe fetal abnormality detected after 13 weeks’ gestation.5 As well, methotrexate is commonly used in the treatment of early, unruptured ectopic pregnancy, in doses well below those used in oncology. The communications I have received on the subject lead me to believe that medical abortion is currently extensively practised in Australia.

Caroline M de Costa

To the Editor: The Australian Fracture Prevention Summit held in 2002 recognised osteoporosis as a major public health issue. Despite this, several Australian studies have found that a majority of patients with osteoporosis-related fractures do not receive appropriate evaluation and treatment as recommended by the clinical guidelines.1-4 In 2003, the Queen Elizabeth Hospital, a tertiary referral hospital which services the north-western suburbs of metropolitan Adelaide, implemented a novel approach to improve the secondary prevention management of patients admitted to the orthopaedic unit with fragility fractures. The strategy was based on the “plan-do-study-action” principle of medical quality improvement, with the primary goal of enhancing performance.5 Before commencing, a retrospective case-note review of 40 consecutive patients who had been admitted to the orthopaedic unit with osteoporotic fractures revealed that, at discharge, none were receiving any medication for osteoporosis. Patients over the age of 50 years who had been admitted with fragility fractures were identified from computer records. With the support of a physician and junior medical staff, a clinical pharmacist provided individual counselling, written materials and osteoporosis therapy. The rate of medication prescription was initially assessed at discharge. In a follow-up telephone interview, participants were queried about the continuation of osteoporosis therapy, performance of investigations by general practitioners, and history of falls. Over a 10-month period, of 259 patients admitted with fragility fractures, 228 patients (88%) were prescribed osteoporosis therapy (calcium, vitamin D and risedronate) on discharge. Of those eligible, 65 patients participated in the follow-up audit. Forty-eight patients (74%) continued to take medications for osteoporosis as initially prescribed; only 28% had had laboratory investigations for osteoporosis and 31% a bone mineral density test. In addition, three patients had experienced recurrent falls complicated by further fragility fractures. The appointment of an allied health champion with clinical backup from a general physician appeared to have achieved a high level of initiation and continuation of osteoporosis pharmacotherapy in at-risk patients during hospital admission. The low rate of follow-up investigations is consistent with previously published data suggesting poor community-based follow-up after hospital discharge of patients admitted for osteoporotic fractures. The major limitation of this clinical pathway is the low rate of patient participation in the follow-up audit. Therefore, the results of follow-up data cannot be said to be representative of the cohort. This study highlights the importance of the participation of GPs in maintaining patient compliance with hospital-initiated programs, especially those involving chronic illnesses.

Tim Yu-Ting Lu · Jennifer A Pink · Lauren E Whitten · Catherine L Hill · Robert J Adams · Catherine Gibb