Out of Order—Grappling With
  Test Use

April 2001
Cover Story

Karen Titus

True story: One morning about seven years ago, Mark S. Lifshitz, MD, got a call from his chemistry supervisor. Seems the lab had just received an unusual order-30 requests for digoxin tests from one nursing unit.

"So I called the resident on the unit to find out what was going on, why there’s so many dig levels all of a sudden," recalls Dr. Lifshitz, director of clinical laboratories at New York University Medical Center. The resident’s reply was a model of brevity, if not logic. "He says, ’Well, we ordered it on everybody because we weren’t sure who was taking the medicine.’"

Here’s another one: Earlier this year, Michele D. Raible, MD, PharmD, received a request for a particular thrombosis test, one she herself was unfamiliar with. "I looked up the test and found out one of our research laboratories here was performing it," says Dr. Raible, assistant professor of pathology at Loyola University Medical Center, Maywood, Ill. When she asked the clinician what he hoped to gain from this early, research-based assay, "He kind of laughed and said he wasn’t familiar with the test either," Dr. Raible says. "Someone had simply mentioned it to him, and he thought it might be interesting to do." When she told the clinician the test would probably not alter patient care and likely could not be billed to the insurer, "He lost interest."

What pathologist doesn’t have a story like these, some barefaced, occasionally droll tale that can be batted about to prove laboratories are indeed importuned by requests for inappropriate or unnecessary tests?

Good as these stories are, however, they’re ultimately little more than that. Producing hard evidence of inappropriate testing is another matter. Using that evidence to change physician-ordering habits is another matter still.

At best, inappropriate testing is a vast and fuzzy proposition, one more often defined by hunches and suspicions than cold facts.

"The perception is always that tests are being overutilized, but it’s difficult to know for sure," Dr. Lifshitz says. "The criteria for when some of these tests are done, under what clinical circumstances, are so varied. And in the lab, you don’t have access to clinical information. There’s a thousand reasons why someone is ordering a CBC. You don’t truly know if they need the CBC, or if they don’t need the CBC."

"Unless you go talk to the clinician and review each chart, it’s hard to say whether a test is inappropriate," adds Eric Hsi, MD, section head of hematopathology at the Cleveland Clinic Foundation.

Here’s how Dr. Hsi describes one situation at his institution. "We recently have had problems with our white blood cell differential orders. Or at least perceived problems," he says. "We started to look into it and found that about 10 percent of our manual differentials were same-day repeats.

"So the question is, Why are they ordering a differential more than once in the same day? Are the results significantly different? Our preliminary answer is ’no.’ Unfortunately, there is little clinical data on what an appropriate time interval between test ordering is."

All he can say for sure at this point, he admits, is that they’re looking into it.

He’s hardly alone in his endeavor. But his journey will be a lonely one, as it is for anyone trying to gain the upper hand on test utilization. It’s the nature of the beast. There’s scant literature on the subject, and even when it’s available, it’s often too finely drawn to be helpful.

"The papers you see are usually targeted to a specific population—like pediatric CBCs in the ER for febrile children," Dr. Hsi says. Such papers are useful to address a particular problem, but reveal little in the way of global ordering patterns. On the other hand, broad amounts of information may be too general to help solve specific problems. And asking other institutions about their ordering practices is a mixed bag, what with the difficulty in matching up size, patient population, and type of facility, not to mention finding one that actually tracks the needed data.

Call it the Goldilocks dilemma—how can laboratories settle on a solution that’s not too broad, not too specific, but just right?

There’s plenty of unnecessary testing going on," says Robert DeCresce, MD, who chairs the Department of Pathology at Rush Medical College, Chicago. "The problem is, it’s all in the eye of the beholder."

The ultimate beholder, of course, is the clinician. Whose not uncommon response to the issue of test use, says NYU’s Dr. Lifshitz, is, "I’m a doctor, I know the patient, I’ll determine what test to order."

"And the fact is, the person is the doctor, and they do know the patient," Dr. Lifshitz continues. "That’s why, for general tests, where the indications are very general, it’s difficult to counter that type of thinking. It ends up being a difference of opinion. And who’s to say who’s right and who’s wrong?"

Harvey W. Kaufman, MD, vice president and chief laboratory officer at Quest Diagnostics, Teterboro, NJ, takes an even stronger stance, saying flat out it’s not the laboratory’s role to determine what is appropriate and what isn’t. "Physicians know when to order a Pap smear. They know when to order a PSA. We make the assumption that doctors who order the tests know what they’re doing."

Does Quest ever see cases to the contrary? "It’s a lot easier to make these determinations in hindsight," he says. "The question of appropriateness has to be asked in real time."

It also needs to be viewed through the eyes of payers and regulators, he insists. The concepts of overuse and necessity have been foisted on laboratories by insurers and Medicare carriers, he says. "The lab is caught downstream, holding onto the specimen and performing tests"-–but not seeing the patient–"while upstream we’ve got groups who are using complex regulations to decrease their spending."

Of course, not even relatively straightforward matters, such as Pap tests, are unimpeachable. An ASCUS Pap test result creates "a significant number of repeat Pap smears and colposcopies and biopsies," says George Birdsong, MD, director of cytology at Grady Health System and associate professor at Emory University School of Medicine, Atlanta. Yet because it’s hard to distinguish the relatively small percentage of women who have meaningful abnormalities from those who don’t, "You could say it’s not overtesting for the women in whom you do find abnormalities," Dr. Birdsong says. The advent of sensitive HPV testing will probably reduce the number of followup tests significantly.

Kandice Kottke-Marchant, MD, PhD, notes a similar ambiguity in coagulation testing. Yes, she says, she does see clinicians take the shotgun approach-ordering everything under the sun and seeing what turns up. The flip side is that for hypercoagulation workups, no one test provides all the answers. "Once you have a patient that qualifies for that testing, multiple tests are appropriate," says Dr. Kottke-Marchant, section head of hemostasis and thrombosis at the Cleveland Clinic Foundation.

Perception also teeters on training and education, not to mention the laboratory’s own performance.

At Grady Health System, Dr. Birdsong reports, the FNA service is frequently asked to do repeat FNAs in patients whose previous tests have been nondiagnostic or negative. "One facet of this is clinician education-the clinicians don’t always realize that on the initial visit that two to four passes have already been made," Dr. Birdsong says.

"But in their defense, we definitely have had instances where the repeat FNA has turned up something that the original FNA didn’t see," he acknowledges. "I’m sure that reinforces their tendency to get repeat FNAs." Moreover, he adds, some clinicians expect zero or near-zero errors from FNAs alone, as opposed to being trained in the triple test approach of looking at the physical exam, FNA, and imaging studies collectively. A clinician whose suspicions lead to a diagnosis in the face of both a negative mammogram and FNA "might be thinking ’The FNAs and mammograms are not helpful in this case,’ while the radiologist and the pathologists are thinking, ’Hey, the triple test worked,’" Dr. Birdsong says.

Dr. Birdsong and his colleagues at Grady have started to include short didactics in their breast conferences. "Not full-blown lectures, but five- to 10-minute talks on a particular topic." The mini-lectures aren’t restricted to pathologists, but are given by everyone involved in breast care, he says, and are proving to be an efficient way to explain the logic behind each specialty’s actions. If they reduce unnecessary testing, so much the better.

Before pathologists can even start chatting with clinicians about test usage, they need to figure out the focus of those talks.

It’s not as obvious as it sounds.

"I couldn’t tell you if we’re ordering the right number of tests," says Dr. DeCresce. "How would I know?"

Say a laboratory gets a request for a thousand CBCs every day, a not unreasonable figure for larger institutions. To identify unnecessary orders, the laboratory would need to comb those thousand requests and determine which ones were ordered twice in a day, then link them to the clinical data, either by visiting the patient on the floor or, if they’ve already been discharged, by tracking down the patient chart. "Imagine trying to do that for a six-month period. It’s impossible," says Dr. Hsi. "Nobody has that kind of manpower to devote somebody to that kind of project, unless they have a grant."

Granted, his institution’s laboratory information system catalogs and reports results, but that’s a far cry from being an open database that can handle complicated search algorithms. Data mining is still beyond the grasp of most, if not all, laboratory information systems, he fears.

At Grady Health System, reports Dr. Birdsong, some tests very likely are overordered due to the coupling of good intentions and the limitations of the current information systems. Grady comprises a 900-bed hospital and clinics throughout the Atlanta area, and it serves a highly transient patient population. A typical scenario, says Dr. Birdsong, would involve a patient who visits a clinic and receives a physical exam, including a Pap test, which turns out negative. "Three months later she’s back for an unrelated problem, and in a quick perusal of the chart, the intern or resident doesn’t find a negative Pap smear within the last year. So they order another one." Grady is in the midst of a major IS upgrade, but until it’s complete, physicians will remain at the mercy of paper charts, shadow charts, and the inability to always lay their hands on a complete medical record.

So where does all this leave the laboratory? Unless the lab sees a bump up in testing, says Dr. Lifshitz, it’s hard to discern poor testing patterns. Since such blips are the exception rather than the rule, "You have to almost follow your hunches about overtesting, and then look and see if you’re right," he says.

Acknowledging that "we can only do so much," Dr. DeCresce and his colleagues at Rush-Presbyterian-St. Luke’s Medical Center have chosen to go after the bigger-ticket items. By eliminating the extensive serum toxicology screens that used to come from the emergency room, for example, the laboratory has been able to save an estimated hundred thousand dollars a year with no ill effect on patient care. The tests were clearly unnecessary, Dr. DeCresce says, noting that when they were being ordered, the results were typically coming back "long after the patient’s problem had been diagnosed and treated."

While he admits high-volume tests will always be a tempting target for review, his preference is to try to tighten control of outsourced tests, "especially specialized genetics testing, which we do not perform, and newer, very high-cost molecular assays coming down the road," he says. "That’s what I plan to keep my eye on."

He stops short of calling these tests unnecessary. "That’s probably the wrong word," he says, offering up another mouthful: "Maybe we should call them low-payoff, low-clinical-value-on-money-expended tests." He compares such tests to certain pharmaceuticals, such as Cox-2 inhibitors, which are valuable to a small percentage of patients yet are being used indiscriminately. Laboratories will face the same challenge, he argues: how to manage highly specialized, expensive, and important tests that nevertheless are meaningless for most patients.

Dr. DeCresce isn’t alone in pursuing the more-bang-for-the-buck approach. At NYU Medical Center, "We’ve tried to look very closely at the low-volume, high-cost tests, those that cost $300, $500, $1,000," says Dr. Lifshitz. "We actually screen every one of those tests." If someone wants to order a PCR test for Epstein-Barr virus, for example, the director of microbiology has to okay the request. Anyone wanting to order an expensive neurologic test has to run it by Dr. Lifshitz.

"We talk to the clinician about the cost of the test, how it’s going to change the clinical approach, and most of the time they back off the test," Dr. Lifshitz reports. In the two years the lab has been doing this, he estimates, it’s saved $30,000 to $40,000 annually.

He’s also had some success in refining the ordering of blood gases and cardiac markers. The efforts succeeded, he says, because they entailed working with a small group of clinicians in a well-defined area-in these cases, the emergency room and ICU physicians. "It’s very difficult to do that in the house as a whole," he says.

"It’s a lot easier to go after defined tests, defined goals, defined locations," he says. "Otherwise you can spend inordinate amounts of time with zero return."

The biggest problem facing Ronald Sirota, MD, chairman of the Department of Pathology at West Suburban Hospital Medical Center, Oak Park, Ill., is repetitive inpatient testing, particularly with house staff. "They order CBCs or routine chemistries more than once, more than twice, sometimes more than three times a day, on patients who don’t need that kind of testing," he complains.

Why the high frequency? Dr. Sirota blames three factors: inexperience on the part of house staff physicians, an assumption that attendings want multiple tests, and convenience.

This is not to say he doesn’t field inappropriate requests for esoteric tests. "I’ve had many battles over that, some of which I’ve won and some of which I’ve lost," he says.

One of his losing battles has been over the performance of cytogenetic testing on all bone marrow specimens obtained to rule out myelodysplasia. Although cytogenetic analysis is appropriate for patients with myelodysplasia, many specimens (probably more than half) that are obtained to rule out this disorder actually do not show it. Cytogenetic testing is not indicated in most of those cases. Since the test costs well over $300, a considerable amount of money is wasted on these tests. Despite these arguments, "We do it when requested," Dr. Sirota admits.

To put an end to what he considers a wasteful practice, "It’s going to take more than me standing up in my own institution and saying, ’This is not right,’" Dr. Sirota says. "It will take literature and evidence stating such testing wastes valuable resources, and it will take a useful algorithm to guide test selection." Until such literature and evidence become available, few options exist for reform. "I realized long ago I was not going to win every argument," Dr. Sirota says.

In a perfect world, pathologists could turn to guidelines that told them which tests should be ordered, and when.

Perfection is nowhere in sight, but that doesn’t mean such guidance is nonexistent. Indeed, though each institution ultimately has to determine its own best course of action, there’s no need to reinvent the wheel each time.

"Even if we can keep up with the literature, we may not have the skills or time to evaluate an avalanche of literature about a particular area," says Loyola’s Dr. Raible. "That’s where we have to rely on our clinical societies or on our hospital groups, because it’s too much for the individual clinician or pathologist." Loyola in particular has found the University Hospital Consortium, a group of around 85 U.S. university medical centers, to be helpful. Its database, says Stephen Kahn, PhD, associate director of clinical laboratories, has helped Loyola identify disease categories that have the potential for laboratory overtesting.

"UHC is a buzz group for us, but it’s not the only one around," says Dr. Kahn, who also chairs the institution’s laboratory cost-reduction group. Large hospital networks can share information about best practices, he says, as can group purchasing organizations and managed care groups.

Solid literature can also bolster the cause, though, as Dr. DeCresce notes, "I usually don’t find too many clinical journals that say why you shouldn’t do a test; most of them talk about new tests that should be done."

Given that clinicians aren’t always abreast of their own literature, "You are free to point it out to them," Dr. Sirota says. Clinical literature is best, he adds-"They tend to believe their own journals more than they believe us"-but when it’s not available, the pathology literature can also serve to persuade.

When both are lacking, Dr. Sirota calls teaching hospitals in his area to solicit ideas for limiting inappropriate tests. (Occasionally, he admits, those whose opinions he seeks disagree with him, "which is a learning experience for me. Nobody has a pipeline to the truth for any of this.")

Thus armed with anecdotes, hunches, data, guidelines, or some combination, how do pathologists go about taming the beast?

"You need an army of people and plenty of time," says Dr. Lifshitz. He’s not joking.

To curb duplicate tests at West Suburban, particularly on noncritically ill patients, Dr. Sirota and his colleagues in the laboratory use a couple approaches. One is plain, old-fashioned talk.

"We talk to the program directors, and make it clear that the attending physicians are responsible for all the tests ordered by the house staff." This fact is no secret, but it never hurts to remind everyone of it regularly, he says. "You can have a short-term hit, everyone’s fine for six months, and then there’s inexorable backsliding," especially as the house staff shifts. "Every July we have a whole new set of people inappropriately ordering tests," he says with a sigh.

With other expensive and wasteful tests, such as repeat stool cultures and in-house HIV testing, Dr. Sirota says, "We demand they have an infectious disease or GI consult before we allow the test to be ordered."

None of West Suburban’s programs to reduce inappropriate testing have come easy, he says. "We always get arguments in the beginning. Physicians don’t want to have restrictions on their autonomy."

Objections can be overcome only with solid evidence and data. "As a lab director, I personally feel I’m on very shaky ground when evidence doesn’t exist, or if it’s just based on my empiric observations," Dr. Sirota says. "So I make sure I have the evidence as well as support from some of my clinicians, and then we go from there."

Laboratories cannot solve problems of overtesting and inappropriate testing in a vacuum, he emphasizes. Without the support of the medical staff and clinicians, "The lab director is powerless," Dr. Sirota says. "This is an area where you’ve got to have some political power. Otherwise, you run the risk of losing your job because you’re not allowing tests to be done."

At NYU, Dr. Lifshitz’s lab has clamped down on overuse of cardiac markers with help from the hospital information system. Up until that point, he says, monthly monitoring of troponin and CK-MB orders revealed that "no one was using any algorithm; they were just ordering everything all the time"—this despite jointly developed laboratory-ER guidelines that were put in place when troponin was introduced.

"So we went back to the emergency room, and we revisited the issue," Dr. Lifshitz says, with just a tinge of "goodfella" gruff to his voice. In addition to talks with ER physicians, the lab had pop-up screens installed on the HIS, which questions physicians when they order multiple tests and reminds them of previous orders and results. This solution has worked, he says, because it doesn’t prevent physicians from ordering tests; rather, it gives them improved information on which to base those orders. "Sometimes they don’t realize a test has already been ordered. Sometimes it’s easier to order something again than to look up a previous result. So if you can provide them that information automatically, it can help reduce orders."

This approach has its limits, naturally. "You can’t do something like that for CBCs or a chem panel-you’d have screens popping up all the time," Dr. Lifshitz says.

Sometimes, says John Coon, MD, PhD, pathologists simply have to throw in the towel. "The tests that people order are influenced not just by the literature, but by individual prejudices," says Dr. Coon, professor of pathology and director of the molecular diagnostics division of Rush Medical Labs, Chicago. "It’s important to realize that if a clinician has no intention of using the data you provide, no matter how sound it is, then it’s worthless. There’s nothing you can do."

But change does come, indeed, has come, say observers, who largely concede inappropriate and overtesting have dropped over the years. (Though they always quickly add, "We’ve still got a long way to go.")

When it comes, however, it comes slowly. Take the matter of PSA testing. While it’s now the norm, the move from prostatic acid phosphatase to PSA didn’t happen overnight. Quest’s Dr. Kaufman suggests that awareness of the PSA test’s availability first reached a significant level thanks to articles published in popular women’s magazines, whose female readership told their husbands about the test; the men, in turn, told their physicians. "But it was a long process," Dr. Kaufman reports. "In the transition most physicians ordered both tests, and it took six to 10 years for them to become comfortable with just the PSA test."

Likewise, he notes, evolving CPT codes can streamline testing, but clinicians cannot be expected to adopt them quickly. "The laboratory has to tolerate and live with the limitations of that process, which is a slow one," he says.

Nor can the laboratory always move swiftly. At Grady Health System, Dr. Birdsong reports, the matter of clinicians requesting repeat FNAs could be improved by standardization, "but it’s a slow process to get there. Everyone has different training and approaches."

Then there’s the matter of practice guidelines and protocols. No matter how thoughtfully developed, no matter how scientifically sound, such guidelines often wend their way into institutions with maddening slowness.

"The example that comes to my mind is troponin," says Loyola’s Dr. Kahn. The literature is inundated with articles demonstrating the value and quality of troponin testing in diagnosing infarction, risk stratification of unstable angina, and now perhaps even guiding interventional therapy, he notes, and in recent years several prominent cardiology groups have published related practice guidelines.

"But it’s still going to take time before this makes it through the physician sector," he laments. "And there will be physicians and cardiologists who, because of their longstanding practices, still think it’s all a lot of bunk."

Even those who are more receptive to new information find themselves hamstrung. "It’s hard for physicians to find the time to get what’s best evidence and try to incorporate that into their practice," Dr. Kahn says.

Dr. Kaufman’s remarks about the lay press and PSA testing can be viewed in another context. Diagnostics companies are taking the previously unheard-of steps to market directly to clinicians as well as to consumers, joining their brethren in the pharmaceutical industry. Can more requests for inappropriate tests be far behind?

Betts Carpenter, MD, PhD, expected that to be the case when Pharmacia Diagnostics approached her about bringing its allergen-specific IgE test onboard at Cabell Huntington Hospital, Huntington, WVa. Pharmacia had recently begun to market its test to primary care physicians, and Dr. Carpenter, who is professor of pathology and vice chair of the Department of Pathology at Marshall University School of Medicine, worried about the potential abuse in ordering.

"I was reluctant," Dr. Carpenter admits. Though the idea of adding routine allergy testing to the laboratory was appealing, she could see the pitfalls as well. Could primary care practitioners order the test appropriately, or would it end up being abused in the same manner as the older RAST tests? Would the lab find itself in a turf battle with allergists?

The laboratory began offering the test at the university-affiliated hospital last October. But in contrast to what Dr. Carpenter feared would happen, "we’re probably having more underutilization than overutilization."

Dr. Carpenter suggests several reasons for the relative quiescence. One, primary care physicians traditionally haven’t been trained to use allergy testing. "So at this point they’re not really comfortable when it’s indicated," Dr. Carpenter says. This may also explain why some allergists at Cabell Huntington have been ordering the test.

It’s also possible the caution with which Dr. Carpenter and her laboratory colleagues helped introduce the test discouraged potential abuse. Dr. Carpenter, along with the laboratory’s supervisor and administrator, met several times with a Pharmacia representative, and the lab investigated the test’s use at a facility in nearby Charleston. The lab also sent out a memo to clinicians and discussed the new test in the laboratory’s monthly newsletter as well as the hospital’s monthly newsletter. In addition, it sent individual letters to members of the medical staff most likely to order the test.

"I always think you send out a letter and nobody reads it, but they did," Dr. Carpenter says. "I was very careful to make the letter short but relevant. If it’s too long, nobody will bother with it. And we sent out the letter at a time when we weren’t sending out a lot of other information from the lab."

Even the best-laid plans can backfire, however. While the addition of allergen-specific IgE has gone relatively smoothly, Dr. Carpenter notes the laboratory’s efforts to improve testing for rheumatoid factor hit some bumps in the road initially, demonstrating the enormous difficulty labs can face in changing test-ordering habits.

Dr. Carpenter did her homework. She identified rheumatologists as the key group to educate, noting that while primary care physicians use rheumatoid factor, they’re more likely to seek out a rheumatologist rather than the lab when they have questions about a questionable test result. She gathered plenty of scientific data to support use of the new test, and for six months the lab split samples for rheumatoid factor testing, comparing patient values between the older, latex agglutination assay and the newer, nephelometric assay. It ran in-house correlations, even though comparative data had been published already, and it produced a correlation that enabled clinicians to match an international unit-based value from the new assay with a range of titers they would have previously gotten on the old test.

That’s not all. Following those studies, Dr. Carpenter sent out a letter to each rheumatologist, in addition to calling them personally to discuss the change. She plied them with the comparison data when they expressed concern, and she assured them the older test would still be available if they felt a problematic specimen warranted it, an offer that didn’t exactly draw raves from the lab. ("The supervisor of the lab was kind of upset with me that I even offered that," Dr. Carpenter admits.) In addition, she sent a letter to every family practitioner.

For all her efforts, Dr. Carpenter initially reaped "some resistance from rheumatology, because this wasn’t how they were trained to do rheumatoid factor. It doesn’t matter how much data you have to show that the new test is better. You’re adding a new level of frustration to their job, and asking them to learn a new test when they haven’t had a problem with the old one."

With the new rheumatoid factor test now in place at Cabell Huntington, "It’s actually gone over okay," says Dr. Carpenter. No one has requested that the old test be run, and complaints have been minimal, "although we may just not be hearing them," she says.

Despite the resistance she encountered, her exhaustive efforts to prepare clinicians for the change seem to have paid off. "It takes a lot more time and effort, but it’s like preventive medicine, in a way. I think it’s much easier to deal with it upfront than to deal with it later on. If you just send out a blind memo saying, ’We’ve changed this,’ that’s when you run into problems."

Once clinicians know that the laboratory won’t try to sneak changes past them, she adds, "They’re going to trust you. Sometimes you have to work harder in some areas, but that makes your work easier in the future. They’re going to be more amenable to your suggestions for changing a test or how it’s ordered."

Her approach is not without risks. "Pathologists sometimes think, ’I don’t want to open a can of worms by calling clinicians and giving them a forum to complain,’" she says. "And that can be true sometimes; you have to be wise in how you do this."

No doubt can remain in pathologists’ minds-curbing inappropriate test use is a thankless and nigh impossible task. So, as the Starbucks barristas are known to say when customers order a double decaf skim latte, why bother?

It may in fact be no more urgent than remodeling a mudroom-hardly worth extensive effort, some say. "When you have taken care of everything else, cut costs in every other way, then you turn to utilization," says Dr. Lifshitz. "Because typically, reducing utilization doesn’t have much impact on staff or equipment-you’re just saving the nominal cost of the test: a few cents here, a couple dollars there. And when you look into changing testing habits, we may even find there’s no overutilization, or perhaps there’s even underutilization."

Others argue the issue cannot be ignored. Inappropriate tests are costly, generate more inappropriate tests, and affect patient care. And ultimately, they say, it’s the pathologist’s job, pure and simple, to help clinicians order the right tests, at the right time, in the right order-which, by default, means pathologists are charged with keeping a lid on test use.

But even those who urge pathologists to fight the good fight agree the good fight is the one that can be won. "Pick your battles," says Dr. DeCresce. "That’s the best we can do." And recognize that some stories are best as stories, and not every one will have a happy ending.

Karen Titus is CAP TODAY contributing editor and co-managing editor.