Letters 496 http://psy.psychiatryonline.org Psychosomatics 46:5, September-October 2005 causal relationship between venlafax- ine and hyponatremia. The precipitous drop in Ms. A’s serum sodium level within 1 week of venlafaxine initiation and a similarly brisk return toward eu- natremia after its discontinuation were associated with the onset and later res- olution of her delirium. When prescribing antidepressants for elderly and medically ill patients, physicians are alerted to the possibil- ity of SSRI- or venlafaxine-associated hyponatremia and to monitor pretreat- ment and posttreatment sodium levels. Patients treated with venlafaxine who have mental status changes need prompt assessment of fluid and electrolyte status. Patients already at risk of the syn- drome of inappropriate antidiuretic hor- mone secretion because of cancer and/ or CNS disease should be managed with particular caution. James A. Bourgeois, O.D., M.D. Sacramento, Calif. References 1. Buff DD, Markowitz S: Hyponatremia in the psychiatric population: a review of diagnostic and management strategies. Psychiatr Ann 2003; 33:318–325 2. Bugunovic OJ, Sotelo J, Madhusoodanan S: Hyponatremia secondary to antidepressants. Psychiatr Ann 2003; 33:333–339 3. Kirby D, Harrigan S, Ames D: Hyponatremia in elderly psychiatric patients treated with selective serotonin reuptake inhibitors and venlafaxine: a retrospective controlled study in an inpatient unit. Int J Geriatr Psychiatry 2002; 17:231–237 4. Boyd IW: Comment: hyponatremia with venlafaxine. Ann Pharmacother 1998; 32:981 5. Masood GR, Karki SD, Patterson WR: Hyponatremia with venlafaxine. Ann Pharmacother 1998; 32:49–50 6. Roxanas MG: Mirtazepine-induced hyponatraemia. Med J Aust 2003; 179:453–454 7.Morton WA, Sonne SC, Verga MA: Venlafaxine: a structurally unique and novel antidepressant. Ann Pharmacother 1995; 29:387–395 Psychiatric Consultations and Length of Hospital Stay TO THE EDITOR: Yasuhiro Kishi, M.D., and colleagues 1 provided an important update by using relatively recent data on the relationship between the timing of psychiatric consultations and the length of hospital stay. Several notes of caution regarding their findings are im- portant and stem from earlier work in this area. 2,3 The relationships among variables remain associations, and it is entirely possibly that unmeasured fac- tors associated with the request for con- sultation might independently be re- lated to the length of stay or that the direction of inference is reversed. That is, certain patterns of clinical need as- sociated with delayed discharge may more likely become apparent later in the hospital stay (e.g., placement prob- lems). Acute medical complications re- quiring psychiatric consultations can occur late as well as early in the hos- pital stay, and when they do, it is likely that a psychiatric consultation (not nec- essarily “delayed”) would uncover problems needing further assessment and intervention. Thus, it may not be entirely correct to assume that these are “delayed consultations” or that these patients experience “poor outcomes” as a result. Nonetheless, the notion of focus- ing to a greater extent on identifying at- risk individuals in the denominator of patients in a hospital rather than pas- sively waiting for an arbitrary numer- ator of psychiatric consultation re- quests to come forward is a wise strategy for improving patient care. Harold Alan Pincus, M.D. Pittsburgh, Pa. References 1. Kishi Y, Meller WH, Kathol RG, Swigart SE: Factors affecting the relationship between the timing of psychiatric consultation and general hospital length of stay. Psychosomatics 2004; 45:470– 476 2. Marcus SE, Pincus HA, Goldman HH, Wallen J: Factors associated with the use of psychiatric consultations in short-term general hospitals. Psychosomatic Med 1987; 49:508–522 3. Wallen J, Pincus HA, Goldman NH, Marcus SE: Psychiatric consultations in short-term general hospitals. Arch Gen Psychiatry 1987; 44:163–168 TO THE EDITOR: With interest, we read the article by Dr. Kishi et al., who re- ported on risk factors for delays in re- ferral and its consequences in terms of length of hospital stay in a large sample of patients referred to consultation- liaison psychiatrists. We compliment the authors for conducting this work because it provides insights into the background characteristics of the popu- lation seen in consultation-liaison psy- chiatry and in the potential limitations of the current referral procedure. To a large extent, Dr. Kishi et al. replicated findings we presented before, i.e., that late referrals are seen in patients with relatively mild forms of psychopathol- ogy, such as no psychiatric diagnosis or adjustment disorder. Also, late referrals were associated with the diagnosis of depression and delirium, whereas sui- cidal ideation and/or behavior and drug-related disorders were more often seen in early referrals. Among the possible explanations for these findings we offered then were the late occurrence of some psychiatric disorders (e.g., delirium) and the rela- tive unobtrusiveness for the staff of some other disorders (e.g., depression and adjustment disorder). In a Euro- pean research group (not a Dutch group, as mentioned by Dr. Kishi et al.), we therefore developed a method to help staff detect patients in need of