Angiotensin-converting enzyme inhibitors (ACEIs) has been recommended by many authorities as a first-line pharmacological option in managing patients with diabetes mellitus and hypertension.^1,2 Indeed, some suggested that diuretics and b-blockers should be used only with caution as the former may aggravate diabetes and the latter may worsen glucose tolerance and mask hypoglycaemia.³ Similar to angiotensin receptor blockers (ARBs, like irbesartan and losartan), the reno-protective effect of ACEIs on halting the progression of diabetic nephropathy and reduction of albuminuria is well pronounced.^4,5 As substantiated by Wright,⁶ meta-analyses show that ACEIs are effective in preserving renal function in both diabetic⁷ and non-diabetic⁸ proteinuric nephropathy (defined as more than 3 g/d).

Indeed, ARBs have been explicitly highlighted by the seventh report of the Joint National Committee⁹ to be effective in reducing progression to macroalbuminuria,^{5, 10} although other classes of antihypertensive agents are also acceptable with firm clinical trial basis.⁹ These include National Kidney Foundation _ American Diabetes Association (NKF-ADA) guidelines,^11,12 United Kingdom Prospective Diabetes Study (UKPDS),¹³ and Antihypertensive and Lipid-Lowering Treatment To Prevent Heart Attack Trial (ALLHAT).¹⁴

Alarmingly, the prevalence of micro- and macro-albuminuria among Asian hypertensive type 2 diabetics have been estimated as 39.8% (39.2-40.5; 95% CI) and 18.8% (18.2-19.3; 95% CI) respectively. Since these figures may imply an "impending pandemic of diabetic cardiovascular and renal diseases in Asia with its potential economic consequences",¹⁵ ACEIs and ARBs seem to be attractive first-line options.

Of interest is the finding from the same study that intakes of diuretics and calcium channel blockers (CCBs) were among the highlighted predictive factors for the presence of macroalbuminuria,¹⁵ hence raising concerns about possible harms of diuretics and CCBs to a known diabetic.

However, recent debate arose as to the best first-line antihypertensive in diabetic patients. For instance, in the Patient-Oriented Evidence that Matters (POEM) session¹⁶ of the British Medical Journal, diuretics was labelled as the least expensive and most effective agent, and "... should be the first line treatment almost everyone with hypertension, including patients with diabetes...". It was argued that none of the other first line drugs were significantly better than low dose diuretics for any outcomes, but that most cost considerably more than diuretics.

In fact, a study of the usage patterns of antidiabetic and antihypertensive drugs in Hong Kong¹⁷ found that most patients with diabetes in the General Out-Patient Clinics (GOPCs) were prescribed indapamide (72.7%) as the only antihypertensive agent, and in the specialist clinics (SCs) ACEIs (35%) and indapamide (22.5%) were the most popular.

Whereas in standard textbooks of family medicine¹ diuretics has been regarded as "contraindications" in prescribing for maturity-onset diabetic patients, recent controversies and study findings may influence patterns of antihypertensive prescription in GOPCs, SCs and integrated clinics.

In response to these inconsistencies in recommendations, the present study investigates how family medicine trainees choose first-line antihypertensives agents in known diabetic patients.

Method

We conducted 101 telephone interviews from December 2003 to January 2004.

The survey enquired on the preferred first-line agent prescribed by the participants in the newly diagnosed hypertensives of different ages and genders with co-existing diabetes (i.e. questions 3 and 4; see appendix which was extracted from the full questionnaire¹⁸). The questions assumed the diabetic patient had no complications on presentation, including various forms of nephropathy.

Eligible participants included the basic and higher family medicine trainees registered in the Hong Kong College of Family Physicians (HKCFP) in the year 2002-2003.

"Basic trainees" are those undergoing hospital- or community-based training offered by HKCFP in the study period. trainees in the college; or had just completed the higher training in 2002 certified and acknowledged by the Board of Vocational Training and Standards (BVTS) in the college's Annual Report 2002-2003.¹⁹

One of the authors contacted each cluster representative (or an equivalent coordinator) for basic trainees, and study group leaders for higher trainees. They were requested to invite trainees' participation. The telephone numbers of trainees could only be obtained when explicit and informed verbal consent of each trainee was available.

After obtaining participants' consent, we further assured interviewees' anonymity, confidentiality and the sole purpose of collecting data for research.

We compared the prescription proportion of ACEIs between basic trainees (including hospital and community-based) and higher trainees using c² tests. We also compared practice experience among those choosing ACEIs versus those not choosing ACEIs by the student's t-tests, with an assumption of normality. In addition, the correlation between physician gender and ACEI prescription was also studied by Fisher's Exact tests. An a value of 0.05 was adopted in all tests, and all comparisons of continuous variables were two-sided.

We defined "higher trainees" as doctors who had completed four years of basic training; current higher

We received 142 telephone numbers, and 101 complete telephone surveys were conducted by a single investigator (Table 1). The other 41 potential participants could not be contacted in the study period. There was no refusal or withdrawal during the survey after thorough explanation of the nature of the interview, as well as the ethical issues concerned.

For the basic trainees, the majority of participants chose ACEIs (67%) as first-line agents regardless of age and gender (Table 2). This was followed by CCBs (16%) in the young patients, whereas diuretics (16%) were more popular in the elderly patients. Overall, patient gender had no effect on physicians' choice of first-line drugs.

Similarly, ACEIs (52% to 64%) were the most commonly prescribed among the higher trainees. Diuretics (32%) also enjoyed higher popularity in the elderly population as compared with the young (16%), in both male and female patients. On the contrary, very few higher trainees chose CCBs.

One higher trainee, who was indifferent to the class of antihypertensives in prescribing for patients without concomitant risk factors, selected ACEIs in the presence of diabetes in all ages and genders. In addition, another higher trainee who chose candesartan cilexetil (Blopress, an Angiotensin A₂-receptor blocker) and bisoprolol fumarate and hydrochlorothiazide (Lodoz, a b-blocker and diuretic combination) for the young and elderly age groups in the absence of diabetes, kept his choice of blopress for the young age group but changed to Losartan and hydrochlorothiazide (Hyzaar, an ARB and diuretic combination) for the elderly if diabetes was present, for both male and females.

The proportion of prescribing ACEIs (52% to 67%) remained high for both basic and higher trainees (Table 3). There was however less higher trainees (52%) choosing ACEIs for the elderly males and females, although the present finding cannot reveal statistical significance.

Analyzing all participants as a whole (Table 4), those who prescribed ACEIs among all patient age and gender groups were found to have lesser number of years after graduation (around 3.7 versus 4.4 years for those choosing other agents). In particular, among the elderly patients, the difference in the number of years was significant (0.02 < p < 0.05) between the "ACEI-prescribers" and "non-ACEI prescribers". In other words, the more experienced trainees tend to choose agents other than ACEIs for the elderly age groups.

As regards the effect of physician genders on prescription pattern (Table 5), more female doctors seemed to prescribe b-blockers than male doctors for both young patients (16.2% vs. 7.8%) and elderly patients (13.5% vs. 3.1%), despite the absence of statistical significance. The other antihypertensive classes did not reveal correlation between physician gender and prescription proportion.

The co-existence of diabetes and hypertension is significant, and they have been implicated as multiplicative risk factors for cardiovascular diseases.²⁰

A thorough discussion of antihypertensive agents in new-onset diabetes,²¹ including trial evidence of Renin-Angiotensin System (RAS) blockade by ACEIs and ARBs, has been reviewed in the December, 2004 issue of The Hong Kong Practitioner. Chan et al²¹ highlighted the limitations of existing hypertension guidelines, and that "drugs blocking the renin-angiotensin system could be considered as initial choice of antihypertensive therapy in overweight hypertensive patients" as one of their key messages.

One might intuitively extend the insightful commentary by Chan et al²¹ to the prescription of first-line antihypertensives patient already diagnosed having diabetes; namely ACEIs or ARBs, could be pretty high in the list of physicians' drug choice. This is especially true as there is still controversy on the best first-line agent. No trainees have chosen ARBs in the present study as first line agents, which may be due to drug availability in their clinics, administrative or cost concerns.

However the choice of first-line agent is still a challenging one as we review more emerging evidence. For instance, the Valsartan Antihypertensive Long-Term Use Evaluation (VALUE) trial²² reminded us of the higher six-month cardiovascular mortality in valsartan-based therapy versus amlodipine-based ones, which could be attributed to poorer blood pressure control of the former. As argued by Cheung²³ based on VALUE, meticulous control of blood pressure should be the prime concern, and that the debate on which class of antihypertensive may only be of financial interest.

The popularity of ACEIs among basic and higher trainees (Table 2 and 3) could be explained by the factors above-mentioned, authoritative textbook teachings,^1,3 and concerns over other agents to wreak havoc (like interstitial nephritis if diuretics were chosen²⁴). The reasons why diuretics were more commonly chosen in the elderly (Table 2) could again be textbook recommendations;¹ whereas the potential sexual dysfunction and electrolyte disturbances accompanying diuretics were rather intolerable in the young, working, sexually active populations who also need regular monitoring of their glycemic status.

The finding that those not choosing ACEIs in the elderly patients were among the more experienced trainees (Table 4) is of interest. This group of physicians could be more exposed to recent evidence- based (and sometimes expert-opinion based) arguments suggesting more evidence of beneficial effects on morbidity and mortality with older antihypertensive drugs, like systematic reviews from MEDLINE and the Cochrane Library²⁵ and the series of 3 articles by Wright.^{6, 26, 27}

Indeed, a recent study by Tam et al²⁸ in the October issue of the Hong Kong Medical Journal, 2004, has shed light into the prescription of ACEIs in different ages and genders. They found a positive association of advanced age (p = 0.002) and female sex (p < 0.001) with microalbuminuria in diabetes in 3 local clinics. Their contributions might indicate a lower threshold for physicians prescribing ACEIs in elderly and/or female diabetes, although we should bear in mind that older people responded less well to AECI suppression, "because their renin-angiotensin systems are more suppressed".²⁹ This speculative assumption should however be explored further by larger scale studies.

In other countries, gender-based differences in the prescription of antihypertensive agents were demonstrated, and were affected by physicians' training level and patient comorbidities. For instance, one study found that male doctors preferred b-blockers while female doctors prescribed methyldopa more extensively in diabetic hypertension.³⁰ Our present study (Table 5) did not reveal the effect of gender on antihypertensive choice in our locality.

We wish to infuse one more important factor to consider in the choice of a first-line agent in the presence of diabetes, namely drug tolerability. Among evidence-based guidelines in recommendations of best first-line agents, this has received relatively little attention as many concerns have been on the cardiovascular end-points. Disregarding the potency of ACEIs or ARBs in halting nephropathy, if diabetic patients are simply not taking the first-line antihypertensives due to their side-effects, or even experience serious adverse reactions causing default to follow up, this would hinder our commitment as primary care physicians in excelling chronic disease management. For instance, would you prescribe captopril to a diabetic if the odds of inducing intolerable cough is as high as, say, 40%?

So far we could barely observe any local data on drug tolerability. Hence we hereby suggest a prospective research studying the tolerability pattern of various antihypertensive agents in the primary care setting, the result of which would then allow a more informative choice of first-line drugs in our locality.

Acknowledgements

The authors would like to thank all participants who have contributed their invaluable time and opinions.

(1)* In your own present practice, which drug would you mostly use as your first-line anti-hypertensive in a young (< 65 y.o.) and an elderly (> 65 y.o.) male patient newly diagnosed as having hypertension (HT)? He enjoyed good past health and had no concomitant cardiovascular risk factors.
在你現時執業的地方，當處理一位剛被診斷為高血壓的 (A) 年輕 (少於65歲)及 (B) 年長 (超過65歲)的男性病人時，你最先會選擇什麼抗血壓藥物？他以往的健康良好，亦沒有其他循環系統的高危因素。

(A) Young HT ____________________________ (Drug name)
(B) Elderly HT____________________________ (Drug name)